Skip to main content

Full text of "Bulletin of the United States Fish Commission"

See other formats


This is a digital copy of a book that was preserved for generations on library shelves before it was carefully scanned by Google as part of a project 

to make the world's books discoverable online. 

It has survived long enough for the copyright to expire and the book to enter the public domain. A public domain book is one that was never subject 

to copyright or whose legal copyright term has expired. Whether a book is in the public domain may vary country to country. Public domain books 

are our gateways to the past, representing a wealth of history, culture and knowledge that's often difficult to discover. 

Marks, notations and other maiginalia present in the original volume will appear in this file - a reminder of this book's long journey from the 

publisher to a library and finally to you. 

Usage guidelines 

Google is proud to partner with libraries to digitize public domain materials and make them widely accessible. Public domain books belong to the 
public and we are merely their custodians. Nevertheless, this work is expensive, so in order to keep providing tliis resource, we liave taken steps to 
prevent abuse by commercial parties, including placing technical restrictions on automated querying. 
We also ask that you: 

+ Make non-commercial use of the files We designed Google Book Search for use by individuals, and we request that you use these files for 
personal, non-commercial purposes. 

+ Refrain fivm automated querying Do not send automated queries of any sort to Google's system: If you are conducting research on machine 
translation, optical character recognition or other areas where access to a large amount of text is helpful, please contact us. We encourage the 
use of public domain materials for these purposes and may be able to help. 

+ Maintain attributionTht GoogXt "watermark" you see on each file is essential for in forming people about this project and helping them find 
additional materials through Google Book Search. Please do not remove it. 

+ Keep it legal Whatever your use, remember that you are responsible for ensuring that what you are doing is legal. Do not assume that just 
because we believe a book is in the public domain for users in the United States, that the work is also in the public domain for users in other 
countries. Whether a book is still in copyright varies from country to country, and we can't offer guidance on whether any specific use of 
any specific book is allowed. Please do not assume that a book's appearance in Google Book Search means it can be used in any manner 
anywhere in the world. Copyright infringement liabili^ can be quite severe. 

About Google Book Search 

Google's mission is to organize the world's information and to make it universally accessible and useful. Google Book Search helps readers 
discover the world's books while helping authors and publishers reach new audiences. You can search through the full text of this book on the web 

at |http: //books .google .com/I 









1 9 O 2 . 

GEORGE M. BOWERS, Commissioner. 

(iovkknmknt i'kintimi office, 

I !♦ O 4 . 


Bcardslcy, A. E. Tht* Dentniction <»f Trout Fry by Hydra. (Isyiie<l Marth 30, lf»03) 155-ltJ<) 

Clark, Hubert Lyman. The PVhirKMkTm.^ of the \Voo<ls Hole R«»gioii. ( Issue*! June 2rt, 19(M ) . 54o-57*5 

Dean, Bashford. Japane**e Oyster Culture. ( Ls8ue<l February 1 7, liH.)3 ) 13-37 

Eigenmann, Carl H. The Fresh-water Fishes of Western Cuf>a. (Issue<l July 20, 11W3) 211-236 

Evermann, Barton Warren, and Jordan, David Starr. Descriptions of New Genera and 

Species of Fishes from the Hawaiian Islands. ( Issue<l April 11, 1^)3 i 161-208 

. Descriptions of a New Genus and Two New Species of Fishes from the Hawaiian 

Islands. ( Issued July 9, 1903) 209-210 

Harron, L. G.,and Smith, Hugh M. Breeding Habits of the Yellow Cat-fish. (I88ue<i March 

30,1903) 149-154 

Hcrrick, C. Judson. The Organ and Sense of Ta.-^te in Fislies. (Issued July 20, 1903) 237-272 

Jenkins, Oliver P. Report on Collections of Fishes made in the Hawaiian Islands, with 

Descriptions of New Species. (Issued July 23, 19t>3) 417-511 

Jennings, H. S. Rotatoria of the l'nite<l States. II. A Monograph of the Rattulidie. ( Issue<i 

Julv23, 1903) 273-^^52 

Jordan, David Starr. Notes on Fishes colle<*te<l in the Tortupaa Archipelagt>. (Issued 

January 19, 1904) 539-544 

, and Evermann, Barton Warren. I)es<Tiptions of New (ienera and Species of Fishes 

from the Hawaiian Islands. ( l8sue<l April 11, 1W3) 161-208 

. I>es<Tiptions of a New (ieuus and Two New Species of Fishes fnnn the 

Hawaiian Islands. (Issued July 9, 1903} 209-210 

, and Starks, Edwin Chapin. List of Fishes dre<ljre<l by the Steamer Albatn»ss off the 

Coast of Ja|>an in the Summer of 19(H), with Desi-riptions of New SixHMes and a Review of 

the Japanese Macrouridji'. ( Issued Aujrust 13, UKM ) 577-6130 

Kendall, William Converse. Habits of Some of the Commercial Cat-lishes. ( Issue^l Aujrust 9, 

1903) .199-109 

. Not**s on Some Fn*sh-water Fishes from Maine. (Issueii August 14, 1903) 35:"J-368 

LydeU, Dwight. The Habits and Culture of the Black Bass. ( Isisued Fel)ruary 26, 190.3 ) . . . 39-44 
Marsh, M. C. A More Complete Description of Bacterium Truttie. (Issue<l Septeml)er 11. 

UK)3) 411-4ir> 

Moenkhaus, William J. Descriptinn of a New Speci(»s of Ihirter from TipjHH*anoe I^ke. 

(Issueil Au^ai.**t 8, VMU) ;J95-;W8 

Parker, G. H. Hearing an<l Allie<i Senses in Fishes, i Issue<l February 26, 190,3) 45-64 

Rutter, Cloudsley. Natural History nf the (^uinnat Salmon. ( I.ssue<l Manh .30, UHKi) 6.V141 

Smith, Hugh M. Ol>st*rvatinns on the Herrinjr Fisherii^ of Fnjrland, Scotland, and Holland. 

1 Issued February 4, U»03 i 1-10 

, ami Harron, L. G. Bree<linjr 1 labits of the Yellow Cat-fish. ( Issue*! March .30, 1903 ) . I4i>-1.>4 

Snow, Julia W. The Plankton Alpi* of I^ike Krie, with S|KH'ial Refen»nce to the Chlo- 

roi>hyceii'. ( Issued August 4, V.m i .' 3«it>-:i94 

Snyder, John Otterbein. A Catalogue of the Shore Fishes colle<*te<l by the Steamer Albatross 

a})Out the Hawaiian Islantls in 1^h:)2. ( Issue<l January 19, 1VK)4 ) olli-o'iH 

Starks, Edwin Chapin, and Jordan, David Starr. List of Fishi»s dre<lgiHl ])y the Steamer 

Albatross off the coast of Japan in the Summer <»f liKX), with Descriptions of New Si>€x.'ies 

and a Review of the Japanese Macrouridie. { Ist-uiMl Augu.<t 1.3, 11M)4 ) 577-6130 



Hrrrino F18HEBIR8 op England, Scotland, and Holland: FrHiij? Phko 

Plate 1. Views of the herring indiwtry of Yarmouth, EuKland 1 

'1. Herring Vi»«iel8 and herring-packing eHtablishment*) at Vhianlingcn. Holland 16 


Plate 3. OyHter park at Tanna. General view showing bamboo collectors arranged in iiarallel lincM '^1 

4. Oysti^r park near Nihojima. Typical living ground, or ike-ba, with hedge of ahibi '2» 

6. Oyster park near Nihojima. General view showing newly arranged toya 3*2 

G. Bamboo oyster collectors, or shibi, after having been in use about one month, A; six months, R; 

eighteen months, C. Detache*! oysters shown at I) St; 

7. Map of the oyster and seaweed c<mcessionK in one of the estuaries of Nihojima to show how com- 

pletely the cultural area is develope<l 3h 

Habits and Culture of the Black Bass: 

Plate 8. Black bass pond, fry retainer, and baw IkhI 40 

Heabino and Allied Senses in Fishes: 

Plate 9. (1) Side view of aquarium, showing sounding apparatus at right-hand end ami suspended glaw cage 
in which the fish were confined. (2) End view of a<iuarium showing sounding apparatus. (3) 
Dorsal view of brain of Fumlultu hrteroclUu*, dis8e<*ted to show iMwitions of the roots of the fifth 
and seventh ner^-es (V), the roots of ninth and tenth nerves (X), and the internal car as Indicated 
by its otolith. (4) Dorsal view of head of Funtlulu* hderocfUiu, to show region where the following 
nerves were cut: The fifth and seventh (V), the eighth (VIII), and the lateral-line nerve (X). (ft) 

Side view of FundtUun hrterodiiun, showing the r^on where the lateral-line nerve was cut (X) 48 

Natural History of the Quinnat Salmon: 

Plate 10. (1) Lower McCloud River Fulls, which prevent the ascent of salmon. (2) Satrramento River at Dun»*- 

muir. (3) Pit River Falls «7 

11. (1) Sacramento River near Sims, "INk)! B," referred to on pag»; 102. (2) Sacramento River in the 

vicinity of Princeton Tl 

12. Diagram showing number and size of young salmon taken at Walnut Grove, isw m 

13. (1) Adult and grilse forms of male salm<m, with genital organs mature. (2) I^nipn^y scar on opercle 

of salmon. (3) Died from gill parasites, last of summer run, September, 1900 120 

14. Diagram showing the passage of two runs of salmon from Vallejo to Sacramento, 1898 124 

15. {A) Female that had spawned all but about 500 ova, showing that the injuries are receivi'd while 

spawning the last few ova or after all have Tmh^u spawiie<l. (B) Male, apparently exhausted fn>m long 
residence in fresh water, but not fn>m l>eing on Hjmwning )>eds; typical condition of late summer- 
run males, Battle Cri>ek, September 15, 1900. ((') Female, with ova liut half develo|ie<1. Battle 
. Creek, September 15, 1900: died from long residence in fresh water. {I), K) Two males, grilsi> and 

adult, showing extreme cases of fungous growth, October 22, 1900 Ki8 

16. Tails of female salmon from spawning beds. {A, B) Tails of spawned-out salmon. {C) Tail (»f bmnde<l 

specimen No. 91, the tail being perfect eight daysbefcire photograph was taken. (/)) Tail of salmon 

with about 500 ova yet remaining 1 40 

17. Sacramento River between Redding and Tehama. Spawning birds of fall Kalmon indicate<l by clusters 

of dots 142 

18. Chart of Sacramento River. Observ'ation stations indicated thus, ( ) 142 

Fresh Water Fishes op Westebn Cuba: 

19. (I) San Juan River, looking upstream from the first bend almve the United Hal>ana Railroail bri<lgi'. 

(2) San Juan River fn>m above the United Hal)ana Railroad bridge 213 

20. (1) Rio San Diego at Paso Real, looking ui)stream from below Western Railnwd bridge. (2> Rio del 

Pinar, looking upstream from bridge 216 

21. (1) Stygicoladentatus(Poey). (2) Lucifuga subtemineus Poey. (2) Lucifuga, a blind fish contain- 

ing unboni young with well-developi'd evi's 23«; 

Rotatoria of the United States: 

Plate I. (1-6) Diurella tigrisMiiller. (7-10) Diurella tenuior Gosse. (11-14) Diurella weberi, new sitccies 862 

II. (15-18) Diurella insignisHerrick. (19-'23) Diun>Ila i»orcellusGosse. (24-26) Diurella sulcata Jennings.. 352 

III. (27-31) Diurella stylata Eyferth. (32-84) Diurella brachyura (tosse. (35. 3(>) Diurella cavia GoKse 352 

IV. (37-39) Diurella rousscdeti Voigt. (40-44) Diurella dixon-nuttalli. new sikk'Ics 352 

V. (45-49) Rattulus gracilis Tessin. (50-52) Rattuluf« s(.'ipio Gosse. (53, 54) Rattuliis macenis Gottse 352 

VI. (56-«») Rattulus multicrinis Kellieott. (69-61) Rattulus capucinus Wient &. Zach 852 

VII. (62-«4) Rattulus cylindricus Imhof. (65, 66) Rattulus latus Jennings 3^)2 



RoTATOKiA OF THE UNITED STATES — Continued. Fa<'in^ page 

Plate VIII. (67-72 > Rattulus lonjrisetA Schrank. (73-76) Rattulus biciutpi^ Pell 352 

IX. (77-W) Rattulus bicristatus (ioe*»e. (81-h6i RattuliL«< piu<illus LauterlMini 352 

X. (^6-91) Rattulus muc<»jnw Stokt-v. (92-94 ) Rattulus stylatus G(»q<* 352 

XI. (95-97) Rattulu-scarinatus Lamarck. (98, 99) Rattuliw loph<K>st«U8 (Jojwi'. (10(i. 101 1 Rattulus rattu> 

Mailer :i:y2 

XII. (102-107j Rattulm> elon^tus Gosse 352 

XIII. (lOK-110) Diurella intermedia Stenroos. (Ill, 112) Katlulus htipio Gosse. tllS) IMurella huU-ata 
Jennings. (114, 115) Diurella brachyura (loese. (116, 117) Diurella weberi, new ppecies. (US, 

119) Diurella sulcata Jennings ;t52 

XIY. (1*20. 121) Diurella RejunctipesG«»sse. (1'22) Diurella helminihodesGosf^e. (1*2:1-126) Diun'lhi marina 
Daday. (127) Diurella collari.s RoiLsselet. (128) Diurella brevidactyla Daday. a29) Rattulus 
cur\-atus Levan<ler. (130j Rattulas brachydactylus Glassrott. (131, 132; "Rattulus lunaris" 

Ehrenberg. (133) Rattulus dubius Lauterlx^m. (134) DIstemma setigerum Ehrenberg 352 

XV. (135) Rattulus unidens Stenroos. (136) RattuliLS ouspidatus Stenro<»s, (137) Rattulus n*seus Sten- 
roos. (138) '• RattuliLS oimolius" Go«se. (139) "Rattulus ral>-ptus" Gosse. (140-143) Elosa wor- 
rallii Lord. (144) " Crelopus (?) minutus" Gosse. (145) B<ithriocen'a affinis Eichwald. (146» 

B<Jthriocerca longicauda Daday 352 

Pla.vkto.v Aui.E OF Lake Erie: 

Plate I. (I) Chlamydomonas gracilis Snow, dl) Chlamydomonas communis Bnow. ^^III) Chlamydomona.s 

globosaSnow. ( IV) Scene<lesmus bijugatuii var. flexuosus Lemm. ( V) Staurogenia apiculata :im 

II. (VI) Fuaola viridisSnow. (VII) Oocystis borgei, (VIII) Choilatella citriformis Snow. (IX»Pleuro- 

coccus regularis .\rtari 394 

III. (X) Pleupoooccus aquaticus Snow. (XIj (*hlon>c<»ocum nutans .Snow. «XII) BotrydiopsLs eriensis 

Snow. (XIII) Botrydiopsis oleacea Snow 394 

IV. (XIV) Chlorfwplijera lacustris Snow. (XV) Chlor(.>spha;ra parvula Snow. (XVI i Mesocari>us spec. 

(XVII) Cu'lospharium roseum Snow. (XVIII) Chrooooccas purpureus Snow 394 

A More Complete Descriptio.v of Bacterium trutt.e: 

Plate I. Bacterium tmtts Marsh 411 

II. Bacterium trutta* Marsh. Pigment pn.Mluce«l in agar cultures 415 

FISHE8 OF Hawaiian Islands: 

Plate I. (1) Dasyatis ha waiiensis Jenkins, new species. (2) Dasyatis wiera Jenkins, new species 420 

II. Gymnothorax thala«soi>terua Jenkins, new .•tpecies 426 

III. Cypeilurus atrisignis Jenkins, new species 436 

IV. Decapterus canonoides Jenkins, new species 442 

Shore Fishf.s of Hawaiia.v Islands: 

Plate I. (It Carchariaa Insularum Snyder, new .specio. y'l • Carcharias nej«iotes Snyder, new si»ecies 53S 

II. (3) Vetemio vern»ns Snyder, new species. (4 ) Sphagebranchus flavicaudiLx Snyder, nrw sjKHMes 538 

III. (5) Callei'helys luteiLs Snyder, new species. 1 6) Aphthalmichthys hawaiiensis Snyder, new specie**. . . 538 

IV. (7) (Jymnothorax nuttingi Snyder, new species. (i<) Gymnothorax benidti Snyder, new si>e<'ies . . .^ . 5:jx 
V. i9) Gymnothorax mucifer Snyder, new siK?<.'ies. (10) (ixTnnothorax xanthiH^tonius Snyder, new 

>'pe<'i»'s ,5:{>i 

VI. 1 11 ) Gymnothorax waialux- Snyder, new .speeies. (12 1 rn)pten.giuN k*u»'urus Snyder, new sijei-ie"* .53K 

VII. (13) Exonaute« gilberti Snyder, new spii-ies .«v3>s 

VIII. (14) Carangus cheili(» Snyder, new .spe<*ies. (15) Caningoide^ ajax Sny<ler. new .•*|M'cies has 

IX. (16) Collybus drachme Snyder, new si>ecies. ( 17 > Ajjogon erythrinus Snyder, new spe<'ies .53x 

X. (IS ) Cirrhilabnis jordani Snyder, new sptH'ies. 1 19) Hemipteronotus .<5nyder. new .specirv . . . 53s 

XI. rJO» Chjit4>don corallieola Snyder, new si>ecies. (21 ) Holacanthus fisheri Snyder, new siKfies ,53k 

XII. (22) Stephanolepis pricei Snyder, new siHH.'ies ."VSx 

XIII. (23) .\ntennariiLs nexilis Snyder, new species. (24 ) Antennarius dueM-us Snyder, new siK"<'ie> ,538 

Fishes Collected in the Tortigas: 

Plate I. (1 ) Ctenogobius tortugsB Jordan, new species. t2) Gnatholepis thompconi Jordan, new sptn-it-s 541 

II. (2) Elacatinus oce&nops Jordan, new species. (3) Ericteis kalishene Jordan, new species SH 


Plate 1. ( 1. 2t .\.*.terias forbesi ( Desor ). ( 3. 4 > Asieria* vulgaris Verrill .Vi2 

2. i5-7i .\sterias tenera Stimps«in. (S. 9 » Asterias austera Verrill .5.>l 

3. (lO.lli Cribrella sanguinolenta ((). F. .Muller». il2,13) Solaster endera ( IlrtziuM .Vi«J 

4. (14.15) A.vterias forbesi (Desor). ^16. 17) A.*iterias vulgari<» Verrill. (IS, 19) .\sterias austera Verrill. 

(20,21) Asterias tenera Stimpson. (22) Cribrella sanguinolenta (<). F. Miiller.. i23) S<)la'it«'r endtM-n 

( Retziu**) .V16 

5. (24-27) Ophiopholis aculeata ( Linnaeus). ( 28-30) Ophiurn brevispnna Say oJSh 

6. (31.32) Ophioglypha robusta (Ayres^. (33,31) Amphipholis S4|namata (Delle Chiajei. (^35,3<h Gor>n»- 

nocephalus aga.ssizii (Stimpson ) 560 

7. (37,38) Ophiura brevispina Say. (39.40) Ophioglypha robusta (.\yresi. (41.42i Ophiopholis aculeata 

(Linnaeus). (43.44) Amphipholis squamata (Delle Chiaje). 1 45-47) Gorgonocephalus aga-ssi/ii 

(Stimpson ) 560 

(4H-52) Arbacia punctulata ( Lamarck) 563 

9. (53-57) StrongyloccntrottLS drobachiensis (( ). F. Miiller » 564 

10. (58-62) EchinarachnJusparma (Lamarck). (63.64) Mcllita pentapora (Gmelim ,5^ 


EcHiNoDERMR OF WooDN HoLP. Re(}Ion— <^ontinued. Facing page 

Plate 11. (65,66) Cuciiinaria fn>iid(N4a ((iniiiicnis). (67) Thyonc briareiiR (Lcmicur). (68.69) ThyoneiiniHemita 
(StimpHon). (70) Ciicumaria pulcherrima (A>tc«). (71) Thyoiic sfabra Verrill. (72) Trochcwtnina 
iM)liticam ( Pourtulos). (73) Caiidina an>iiata (Gould). (74) Synapta inhaereiis (O. F. Muller). (75) 

SyimptH roRoola ( Verrill ) 566 

12. (76-80) Curuinuria froiulowa ((tunneriw). (81-85) Cu<:umaria pulcherrima (Ay res). (86-W) Thyonc 

uniHe.mita (StimpHon ) 566 

13. (91-W) Thyone Hcabra Verrill. (95-102) Thyoiie brlnreus (Lesueur). (103,104) Caiidina aicnata 

((lOiild ). (105-108) Troehostoma ooliticum ( PourtaKV) 56K 

14. (109-112) Syiiaptalnha.ren8(0. F. Muller). (113-116) Synapta roseola (Verrill) 571 

Japanese Fikheh Collected by the Albatross: 

Plate 1. (1) Chlorophthalmus albatrosris Jordan & Starks, new Hpecici«. (2) Chauliodus emmelaH Jonlau 

«t Starks, new spofies r»30 

2. (1, 2) NeoacopeliLS alcocki Jordan & Stark?, new specicf*. Polylpnus atcreope Jordan & Stark**, new 

KpecieH 6.10 

3. (1, 2) Pcristedlon amlscus Jordan & Starks. new f»peck»». (3) Watanea fiivicola Jordan & Snyder 630 

4. (1) Melanobranchuis antrodes Jordan «& GiU>ert. new itpccics. (2) Nezumia condylura Jordan & 

Gilbert, new species 630 

5. (1) Athercsthes everinanni Jordan & Stark.o, new species. (2) Alaeops plinthuH Jordan & Starks, new 

species 630 

6. (1) Dexistes rikuzenius Jordan & Starks, new species. (2) Araias ariommua Jordan & Starks, new 

species 630 

7. (1) Versequa achne Jordan & Starks, new species. (2) Microstomus kitaharse Jordan & Starks, new 

species 630 

8. (1) Engyproflopon iijimie Jordan <fe Starks, new species. (2) Scaeops grandisquama (Schlegel) 630 


Japanese 'Oyster Culture: I'age. 

Map of region of oyster culture on north shore of inland sea near Hiroshima (Sea of Aki) 20 

Fisherwoman opening oysters 21 

Hand pick for making sockets in gravelly bottom 22 

Bamboo collectors (shibi ) forming boundary hedge 22 

Arrangement of branched collectors 23 

Diagrams of oyster farms 24, 25. 26, 29, 30, 34, 35 

Oyster hook 27 

Oyster rakes 27 

Mitten used to hold oyster-bearing shibi while seiMniting oysters 28 

Basket for collecting and storing marketable oysters 28 

Bamboo collectors as arranged in Kiisatsu 81 

Bamb<x> collectors as arranged in Kaida Bay 32 

Ground plan of a mound toya of collectors 83 

Natural History of thk Quinnat Salmon: 

Stomach, pyloric appendages and part of the intestine 126, 127, 128 

Heads of salmon, showing changes in fresh water 130, 131 

Genital organs of hermaphrodite salmon 132 

Spawned-out female 138 

Fishes kron Northeastern California: 

Pantosteus lahonton Rutter, new species 148 

AgoHia robusta Rutter, new species 148 

Fresh- water Fishes of Western Cuba: 

Fundulus cubensis Eigenmann, new species 228 

(jlaridichthys falcatus Eigenmann, new sjiecies 224 

Glaridichthys torralbasi Eigenmann, new siKH'ies 225 

Girardinus garmani Eigenmann, new species 226 

Toxus riddlei Eigenmann, new species 227 

Heterandria cubensis Eigenmann, new species 229 

Atherina evermanni Eigenmann, new species 229 

Eucinostomus mei^ki Eigenmalfn, new species 229 

Heros tetracanthus torralbasi Eigenmann, new subspecies 230 

Heros tetracanthus tetnu'jinthus Cnvier and Valenciennes 231 

Heros tetracanthus griseus Eigenmann, new subspecies 232 

Heros tetracanthus latus Eigenmann, new subspecies 238 

Heros tetracanthus cinctus Eigenmann, new subspecies 284 

Heros nigricans Eigenmann, new species 235 


Thk Organ and Sense of Taste in Fishes: Page. 

Brain of yellow cat-fish 242 

Section through skin of top of head of Ameiunis mela.s 248 

Projection of cutaneous branches of communis root of facial nerve in Ameiurus melas 249 

Rotatoria of the United States: 

Dorsal views of toes in Rattulidie 2H4, 286 

Spiral path followed by Diurella tigris Miiller 296 

Diagram of a reaction to a stimulus in Diurella tigris Miiller 297 

Notes on Some Fresh-water Fishes from Maine: 

Leuciscus carletoni Kendall, new sj)ecies 358 

Pimephales anuli Kendall, new si>ecies 360 

Coregonus labradoricus Rieharrlsion 304 

Coregonus quadrilateralis :W5 

Coregonus stanleyi Kendall, new species 367 

Description of a New Species of Darter from Tippecanoe Lake: 

Hadropterus evermaiini Moenkhaus, new species 398 

Report on Collections of Fishrs made in the Hawaiian Islands: 

Congrellus bowersi Jenkins, new 6p<x'ies 422 

Microdonophis macgregori Jenkins, new spei'ies 423 

Mursena lampra Jenkins, new spc-cies 423 

Mursena kauila Jenkins, new species 424 

Gymuothorax leucostictus Jenkins, new species 425 

G>'mDothorax gracilicauda Jenkins, new species 426 

Gymnothorax leucacme Jenkins, new species 427 

Gymuothorax ercodes Jenkins, new spwcies 428 

Echidna leihala Jenkins, new species 429 

Echidna vincta Jenkins, new species 430 

Echidna obscura Jenkins, new species 430 

Echidna psalion Jenkins, new species 431 

Myripristis sealei Jenkins, new species 439 

Seriola spama Jenkins, new species 442 

Carangus hippoides Jenkins, new species 443 

Carangus rhabdotus Jenkins, new species 445 

CarangiLs politus Jenkins, new sx)ecies 446 

Fowleria brachygrammus Jenkins, new sp)ecies 448 

Ap(^on menesemus Jenkins, new spwcies. 449 

Priacanthus meeki Jenkins, new species 451 

Eteliscus marshi Jenkins, new sj)ecies 452 

Pseudupeneus porphyreus Jenkins, new spjecies 455 

Chromis elaphnis Jenkins, new spe<*ies 457 

Calotomus cyclurus Jenkins, new species 4f4> 

Calotomus snyderi Jenkins, new species 467 

Scaridea zonarcha Jenkins, new species « 468 

Scaridea balia Jenkins, new species 469 

Teuthis leucopareius Jenkins, new species 476 

Teuthls gruntheri Jenkins, new species 478 

Acanthurus incipiens Jenkins, new species 480 

Callicanthus mctopos<iphnm Jenkins, new ^|HH.Mes 4H2 

Tropidichthys oahuensis Jenkins, new species 485 

Tropldichthys cpilampnis Jenkins, new species 4S6 

Laetoria galeodon Jenkins, new species 4hH 

Diodon nudifrons Jenkins, new species 488 

('irrhitoidea bimacula Jenkins, new species 490 

Sebastopsis kelloggi Jenkins, new si)ecies 493 

Sebastapistes corallicola Jenkins, new species 494 

Seba8tapb»tes coniort^i Jenkins, new species 49G 

Sebastapistes galactacma Jenkins, new species 497 

Dendrochirus chloreiw Jenkins, new species 498 

Kviota epiphanes Jenkins, new species 501 

Chlamydes laticeps Jenkins, new species !nx\ 

Gobionellus lonchotus Jenkins, new species rM 

Enypnias oligolepis Jenkins, new species ,'i04 

Tripterigion atriceps Jenkins, new six'cies 505 

Salarias cypho Jenkins, new si>ecie8 507 

Salarias saltans Jenkins, new species 508 

Kalarias rutilus Jenkins, new species .'1O9 

Aspidontus brunneolus Jenkins, new species 5io 


Japanese Fishes Dredged by Albatrom: Page. 

Myxine Karmani Jordan & Snyder 677 

Centroscylliiim ritU*ri Jordan & Fowler 678 

Raja tcngu Jordan & Fowler 678 

Diaphus watasei Jordan & Starks, new specieH 681 

Synaphobranchufl Jenkins! Jordan & Snyder 6ft2 

C)ongrellu8 megastomiw (Gltnther) f»2 

Sphagebranchus moseri Jordan & Snyder 6K2 

Macrorhamphofiiui sagifne Jordan & Starkit 68S 

Hippocampus sindonis Jordan & Snyder 6K3 

Paratrachichthys pro8themiuM Jordan & Fowler 684 

Zen ilea Jordan & Fowler 684 

Apogon lineatus Schlegel 686 

Antigonia rubeat'ens (GQnther) 6H6 

Thysanichthys crofwotiis Jordan & Starkn 687 

Lythrichthys eulaben Jordan & Starkn 688 

Scorpeena izenitis Jordan & Starks 689 

()co8ia vcHpa Jordan & Starks 680 

Stlengis OHeasia Jordan & Starkn 590 

Schmidtina minakia Jordan & StarkH 690 

Daruma sagamia Jonlan A, Starks 690 

Ricuzenius pinetonim Jordan & StarkH 691 

PseudoblenniuK totomiiH Jordan & Starks 691 

Cottiusirulus schmidti Jordan «fc Starks 692 

Cottunculus brephocephaliis Jordan «fe Starks 692 

Crystallias matsushimie Jordan & Snyder 692 

Peristedion orientale S<'hlegel 698 

Lepldotrigla abyKsalis Jordan «fe Starks, new si>et'ies 696 

Lepidotrigla Japonica ( Bleeker) 696 

Suniga fnndicola J(»nlan & Snyder 697 

Chteturichthyn fK'iistius Jordan & Snyder 6*7 

Trypauchen wakiu Jordan & Snyder 697 

Callionymus flagris Jonlan & Fowler 698 

Draxronetta xenica Jonian & Fowler 698 

Pteropsaron cvolans Jordan & Snyder 699 

Osopearon vere<uindii« (Jordan & Snyder) 600 

EulophlaH tanner! Smith 600 

Lycenchelys popciiimon Jordan & Fowler 600 

Bothrocara zesta Jordan & Fowler 601 

Pon^aduK guntheri Jordan & Fowler 601 

GadomiLM oolletti Jordan & Gilbert, new specieu 604 

Regania nipponica Jordan & Gilbert 606 

('oryphaMioIdes awie Jordan & Gilbert, new species 609 

CkjryphBpnoides garmani Jordan A, Gilbert, new Rpe<*ies 610 

C-oryphajnoides misakius Jordan & Gilbert, new spetriea 611 

Hymenocephalus striatissimus Jordan & Gilbert, new species 613 

Ilymenocephalus papyraceus Jordan & Gilbert, new species 614 

H>'mcnocephalu8 lethonemus Jordan & Gilbert, new species 616 

( V]elorhynchiis kishinouyei Jordan & Snyder 618 

(kelorhynchus anatinwtris Jordan & (JillK'rt, new si)eci€>s 619 

('lei8theni»s pinetonnu Jordan & Starks, new si>ecies 022 

Lophins lltulon (Jordan) 627 

Malthopsis tiarella Jordan 628 

i. •- . H «;A 



> • 




^^^Bh^UC" — ■^^"miriSISM 




.-«^' '■'' 


vrews OF the h 




Tlie herring {Chipea haremjus) has justly l>eiMi called the *'king of fishes." 
Although its importance is now relatively less than it was several centuries ago, it is 
to-day a leading fish in the Unite<l States, Canada, Newfoundland, England, Scotland, 
Ireland, Holland, France, Norway, Sweden, and Russia. A si>ecies verA' similar to 
that of the Atlantic Ocean is found in the North Pacific (Jcean, and is caught in 
large quantities in Japan and Alaska. In point of number of individual fish taken 
for market, no species exceeds the herring. The annual value of the herring fish- 
eries is about ?525,250,0(X), representing 1,5(XJ,(MKJ,<K)0 pounds of fish. 

In 19(X) the \^Titer visited the principal herring-fishing centers of England, Scot- 
land, and Holland. The following notes, based on the observations then made and 
the information there collected, are presented chiefly l>ecause of the large consump- 
tion of European herrings in the Unit^ States and because of the desirability of 
applying the foreign fishing and preserving methods to the herring industry on our 
east and west coasts. No attempt is made herein to furnish a complet/O account of 
the herring fisheries of the countries mentioned. 

The writer was very courteously assisted in his inquiries by the following jiersons, 
to whom special acknowledgments are due: Mr. Charles E. Fryer, London, one of 
the three Government inspectoi-s having jurisdiction over the fisheries of England 
and Wales; Mr. W. C. Kol>ertson, Edinburgh, secretar}' of tlie fishery Iwai-d for 
Scotland; Mr. J. K. Nutman, of Great Yarmouth, England; Mr. James Ingram, Gov- 
ernment fishery officer, Al)erdeen, Scotland; Mr. E. A. Man, United States consular 
agent at Schiedam, Holland, and Messrs. C. Van der Burg & Son, Vlaardingen, 

Although the capture of herring is already one of the leading fisheries of the 
United States, the writer l>elieves that the industry may l>e increase<l and the trade 
made more profitable by the adoption of foreign methods with a view (1) to supply 
from local fisheries the very large quantities of pickled herring now imi>orted from 
Europe and Canada, and (2) to open a large trade with southern Euroi)e and other 

The following letter from Mr. F. F. Dimick, secretary of the Boston Fish 
Bureau, dated April 7, 1900, is i)ertineut to this subject: 

The herring imported from Norway, Holland, and Scotland are of a different <inality from 
the herrinj? found on <>nr coast. They are fatter, and grejit care is taken of the fish when caught 
and in packing them. The herring caught on our coast oi the same size are not so fat. Our 
fishermen generally find a good demand for their herring at from §1.50 to $3.50 per barrel fresh for 

F. c. B. 19(W— X 1 


bait for the cod and haddock fisherman. The packers generally receive from $8 to $4.50 per barrel 
for United States shore herring, and there is generally not enough to snpply the demand. The 
foreign herrings are consumed principally by foreigners, and sell at from $8 to $14 per barrel. 

The quantities of pickled herring impoi*te(l into the Unitecl Stnt-es in 1900 from 
the countries stat-ed were as follows: 

CV>un tries. 

PoundH. Value. 



United Kingdom naH^USi $12,«>43 


09.12:3 ; 8, (Ml 




$375, .'Wl 



Grimsby and Yarmouth, the important English fishing centers on the shore of 
the North Sea, are extensively engaged in the herring fishery. At Grimsby the 
herring fishery is overshadowed by the beam-trawl fishery for l>ottom fish, but at 
Yarmouth the herring fishery predominates. The great herring markets of Scotland 
are Aberdeen, Fraserborough, and Peterhead, also on the North Si»a. Whih? these 
notes are based primarily on visits to Yarmouth and AlK^nleen, supplementary 
information was gathered in London, Edinburgh, and other plaices. Although the 
same general methods are jidopted in the herring industry in different parts of 
each country, it must be remembered that the accompanying notes are espi»cially 
applicable to the particular places mentioned. 

The i)rincipal fishing season at Yarmouth is from some time in October to the 
last of December. A few fish are (*aught earlier in si)ecial net«; these are small, 
well-flavoi-ed fish known as "hmgshore herring," and are for local consumption. 
There is also a spring fishery, involving most of the month of March and lasting five 
or six weeks. The fish then caught are small and i)oor, and are used for bait in the 
line fisheries. 

The Yarmouth herring fishing is carried on with sailing vessels calUnl luggers, 
and also by steamers. The use of the latter is increasing, 50 to 70 being operated in 
1900. The crew of each kind of vessel consists of 11 men. Each vessel carries 200 
gill nets, 30 yards long, 20 yards (or 200 meshes) deep, and costing £2 apie<*(\ The 
number of meshes to a yard is from 28 to 31, the average size of mesh, bar measure, 
being equal to a shilling. The net« are cotton, machine- mmle, and, with proi>er 
care, may last seven years. They are first tanned with a solution of hot *'kutch" 
or catechu from Burma, which is said to \>e better for this pur^Kjse than tnu bark. 
After drying they are thoroughly soaked in linseed oil, and again dricMl by spreading 
on the ground. Oiling is done only once, but soaking in the astringent solution is 
rejwated from time to time. 

The entire complement of nets is shot at one time, the nets being tied together. 
A duplicate set of nets is held in reserve. The water is thick for 15 or 20 miles off 
Yarmouth, and fishing may be done at any hour of the day or night, but the l)est 
times are about sunrise and sunset. The herring do not remain long in the nets 
unless storms prevent hauling. Some vessels run fish in fresh, othei*s dry-salt their 


catch at sea, and store it in the hold in compartments. A vessi*! ma}' leave port, 
set nets, make a catoh, and be back the same day, or it may be out two weeks. 

There has never been a failure of the Yarmouth herring fishery; although fish 
are less abundant some years, they have never been so scarce as to make fishing 
unprofitable. Some of the Yarmouth fishermen think that the herring fivquenting 
that part of the English coast constitute a distinct lK)dy which spawn and remain off 
that coast, and do not come down from the North. Herring are taken much earlier 
in Scotland and northern England than hei'e. 

When a vessel arrives in poi-t the fish are lifted out of the hold in baskets and 
spread on deck, where they are counteil into baskets by hand, 100 fish to a basket. 
These baskets are then passed over the rail to the dock and emptied into large, 
peculiarly shajwd baskets holding 5(M) fish, arranged on the dock in lines or tiers of 
20 baskets each. The fish are heaped in 10 piles over the edges of adjoining baskets 
to facilitate counting. A line of the large baskets constitutes a last, which is the 
unit of measure in the herring trade. A last represents about H tons of herring, 
or, theoretically, 10,000 fish; but, as a matter of fact, 13,200 fish of any size, as 132 
fish are called ICK) in counting. 

Herring are sold at public auction by la^sts. The buyer puts his card or t-ag on 
the first basket of the tier, and his drayman comes shortly afterwaiils and takes the 
fish to the pickling-house or smoke-house. Sometimes, at the height of the fishery, 
1,(X)0 lasts (or 3,0<X),(XK) pounds) are landed and sold daily in Yarmouth, and the 
wharves i)resent scenes of great activity and excitement. 

The Scotch herring fishery is rather uncertain. In 1900 it was ix)or on the 
Scot<»h coasts, except at)out the Shetland Islands, where there was a phenomenally 
large run. A number of years ago, after expi?nsive curing establishments had l)een 
built in those islands, the fish disapi)eared, fishing had to be abandoned, and the 
packers lost all they had invested. 

Different races of herring are recognized as frequenting different parts of the 
Scotch seaboai*d. Thus, according to Mr. W. C Robertson, of the fishery l)oai*d, the 
l)e8t herring ai-e those taken near Barra and Loch Fyne, on the west coast. These 
are fine, fat fish, which have brought as much as £(> per bari-el. 

The different kinds of cuivd herring to which reference may l)e made are 
ordinary pickled fish, kipi>ered herring, bloaters, and red herring. 


A favorite form of preser\'ed herring for local consumption is the bloater. In 
the United States this term has come to mean a large, lightly smoked herring, but 
in Great Britain a herring of any size may be a bloater, which may be defined 
as a round herring, lightly salted and lightly smoked, and intended for immediate 

The extensive trade in Yarmouth bloaters which formerly existed with London 
and other cities away from the coast has to a great extent died out, owing to the 
fact that the smoking is now done at the place of consumption, llie fish l)ear 
the rail shipment better before smoking than after, so that the bloater trade now 
consists largely in shipping lightly salted fish to cities where there are smoke-houses. 
Bloatei-s remain in good condition for two or three days, but are regarded as being 
best when smoked and eaten the day after being caught. 


The essential steps in the preparation of bloater herring are as follows: Imme- 
diat^ely after being caught the fish are dry- sal ted from 12 to 24 houi-s if fat, or only G 
hours if lean. They are then smoked for 4 to 16 houi*s and are ready for consumption. 

Yarmouth bloaters bring a good prie^*; sometimes as much as 17.s. 6rZ. is 
received for 1(X) fish. 


A special grade of salted and smoked herring is known to the English and 
Scotch trade under this name. The fish are destined chiefly for the Italian, Grecian, 
and general Mediterranean trade, but some are sold in London and other parts of 
Great Britain. Some of the herring dealers handle only red herring; but, as a rule, 
the prei)aration of red herrings is incidental to the packing of other grmles. 

The fish which are destined to be made into red herrings are often those which 
have been kept at sea for several days to a fortnight, and hence have become too 
hard, from prolonged salting, to be made into bloatere, kippers, or regular pickled 
herring. If they have been salted too long on the vessels they are spitted on sticks 
and softened by steeping them in fresh water. The specual peculiarities of red 
herring are that they are round, are rather heavily salted, and are smoked for a long 
time to give them a good rich color. 

When intended for export to very warm countries red herring are salted for 30 
to 48 hours in strong brine and are then smoked for a fortnight or thi-ee weeks. 
For temperate or cold (countries the fish are kept for a shorter time in pi(*kle and are 
smoked 10 or 11 days. Hard-wood sawdust and hard-wood sticks are considered 
necessary in producing the smudge and heat required to give to these fish their 
peculiar flavor. 

Red herring for the MeditiCrranean trade are packed with tlieir heads against 
the barrel and their tails at the center, in dry-ware barn*ls holding 500 to 000 fish, 
half barrels holding 300 to 350 fish, and k(»gs or third barrels holding 180 to 2(K) fish. 
The average gross prices for these packages in recent years have been lO.s., 5.s. to O5., 
and 3j6*. to 4s., respectively. The expenses on a barrel for freight and commissions 
are about 3s., the fish being sent by rail to Liverpool and thence by water to the 
Mediterranean. They are sometimes packed in tin cases when destined for especially 
warm countries, and for the London market they are packed in flat boxes holding 50 
to 00 fish. 


Among the various kinds of prepared herring none ranks higher than kippered 
herring. The essential characteristic of kippered herring (and of all kippei-ed 
fish) is that, before being salted and smoked, they are split and eviscerated. Fish 
intended to be made into kippers should be very fresh whcm received from the 
vessels. At Yarmouth large fish are preferred for this method, while at Aberdeen 
small, fat fish are prefeiTed. 

As soon as received they are split down the back from tail to head, eviscerated, 
and then salted in strong brine of Liverpool salt for 15 to 00 minutes, according to 
fatness. They are then spread on square sticks by means of hooks, an<l smoked 
over a hot fire of hard-wooil shavings for to 8 hours (Aberdeen) or 10 to 10 hours 
(Yarmouth), requiring constant attention. The color imparted to the skin is either 
golden or light, to suit the markets. After cooling they are packed in boxes 


holding 4 to dozen, and are sold at good prices throughout the United Kingdom. 
They have longer life than bloaters, will easily keep for three to ^vg days, and in 
cool weather a fortnight, but they should be eaten as soon as possible. 


Under this name are officially recognized the herring brine-salted and packed in 
barrels and half barrels. Such fish ai-e more extensively prepared than all others 
combined, and give to the English and Scotch herring trades the importance they 
have attained. 

Various grades of herring are recognized by the salt-herring trade. These 
grades are sharpl\' distinguishe<l and are usually indicated on the outside of the 
ban'el l)y a brand, branding is more generally practiced in Scotland than elsewhere. 

The grades of salt herring in England are "mattie," *'mat full," "full,'' and 
*'spent" or *\slu)tten." "Matties'' are the smallest herring, 8} to 0^ inches long, with 
undeveloptHl reproductive organs; "nuit fulls"" are fish U\ to Wi inches long, with 
the ovaries and spermaries h'ft in; "fulls" are fish 10| inches or more in length, 
with roe or milt; and "si)ent" fish are at least 10:j^ inches long, with eggs or milt 

The grades as recognized in Scotland are "mattie," "mat full," "full," "large 
full," and ''spent," and several other minor grades. The word "mattie" originally 
meant a maiden herring, "mattie" being th<» terminology of the east coast and 
"matje"" of the west coast. "Matje" still retains the original meaning, the herring 
so designated Ix^ing caught in May and June; all such fish when salted are now 
sent to Russia. "Mattie," however, represents a small herring, full of either roe 
or milt, or even spent. The official requirements of the hen'iug of the various 
grades are as follows: "Mattie," not less than IMnches long; "mat full," not less 
than U^ inches long, with roe or milt well developed and clearly seen at throat; 
"full," not less than KH inches long, with roe or milt; "large full," not less than 
114 inches long, with roe or milt; "siHMit," not less than 10^ inches long, without 
roe or milt. 

The lengths of salted herring i^pecified under the difTerent grades apply to the 
fish after shrinking, and are measured from the end of the snout t^) the end of the 
compressed tail fin. SiK*cial measuring sticks or gages aiv employe<l by the fishery 

The ccmtinental markets require fish that are gilled and gutted but not split. 
Herring are gutted through the gill cavity, the heart, liver, and repnxluctive organs 
being left /// sifu, but the greater part of the stomach and intestines IxMng removed. 

Gutting is done as soon as the fish are landed, by a civw of three women, two 
of whom do the gutting while the third first " rouses" the fish (i. e., stirs them by 
hand in "rousing"" tubs of water to remove dii-t, blood, etc.) and then packs them 
in barrels with the proper amount of salt. In gutting, a small knife is inserted 
through the isthmus and, with the forefinger or thumb, draws out the viscera. The 
roes and spermaries are always left in, as they ai*e considered food delicacies and, 
in addition, give the fish a fat or plump ap|>earance. Sometimes the roes are so 
large that in packing they rupture the abdominal wall. Although excellent fish, 
they can not, in this condition, l>e sold as the best grade. 


For rousing, Liverpool salt is used; but for packing, eoai'sor Spanish salt is 
employed, about 100 pounds of salt being required for each barn4 of fish. 

In packing herring, it is customary to pack 7 barrels with a ton of fish (l\100 
I^ounds), there Ixjing 300 pounds in a barrel. Each barrel contains fmm 850 to 1,100 
fish, according to size. In packing, each herring is carefully arranged in a definite 
position, with the abdomen upward and with the head against the side of the barrel, 
the fivsh in a given layer or tier l)eing parallel. The fish in the next layer are 
arranged in the same way, but their long axis is at right angles to that of the fish 
in the adjoining layer. The barrels are filled with alU^rnatc? layers of fish and salt, 
and then headed. In packing, the fish are compressed vertically and their l)ellieR 
are flattened, giving them the appearance of being largt^r and rounder. Laterally 
compressed fish an* not in demand. 

During the proc^ess of curing, the fish shrink considerably and the barrels haye 
to bo refilled. In Scx)tland the law requires that the barrels rest on their side and 
be refilled after 11 days. In England, where there is no law, about 8 days are 
aUowed to elapse. A bunghole is lM)red 13 inches alK)ve the bottom, th(» ban-el is 
placed on end, the head is removed, and the pickle is allowed to run off; then the 
hole is closed, 2 to 5 tiers of fish of the same catch are placed on top, and the 
barrel is closed, placed on its side, and the original pickle is returned through 
the bunghole. No new pickle is introduced, and under no circumstances are the 
fish washed in water. After branding, the barrel is ready for market. A well- 
cured and well-packed barrel, after the lapse of 10 full days, should contain no more 
undissolved salt than would fill a cylindrical tub 9 by 1) inches. 

The prices of salt herring vary greatly, depending on the supply. The average 
price of the bc^st grades is usually about 30.s'., but it may drop to 20.s. or rise to 40.*?. 
In 1899 the prices in the Gernmn cities of St^^ttin, Konigsburg, and Danzig, and also 
in Russia, were the l>est ever known, "matties" bringing 24 to 34 marks i)er barrel, 
*'mat fulls" 32 to 30 marks, and "fulls" 30 or more marks. From thest^ gross 
prices, exi>enses amounting to alwut 4^ marks per barrel were deducted. In 1890-97, 
when there was a large catch in Scotland, the average prices of salt herring in 
(xermany were 13 or 14 marks for "matties," 10 or 17 marks for ** mat fulls," and 22 
or 23 marks for "fulls." 

The authorities and fishermen of Scotland fully appreciate the importance of 
plainly designating on the barrels the quality of salted herring, and the fishery boanl 
has formulated a very complete system of regulations governing branding. In view 
of the benefits which have accrued t^) the Scotch herring fishery from the operation 
of the branding regulations, and l)ecause of the importance with which th<^ ])resent 
writer regards branding Jis applie<l to the lTnite<l Stakes herring trade, the following 
detailed refei*en<*es to the subject, are nuwle. 

The official branding of barrels of salted herring is not (*ompulsorv, and only 
alK>ut half the packers resort to branding, but it is generally n'garded as facilitating 
the sale of fish. A good judge of herring would be able, from perscmal inspection, 
to buy just as good fish without the brand as with it; but in distant markets the 
brand carries a guaranty. The fee charge<l by the governm(»nt for aflixing the 
official brand, certifying to the quality of the fish, is 4f/. (10 cents) per bari'el. 
During the years 1898 and 1899 the fees from this souive aggr<»gatcd €11,500, or 
about »57,500. 


The following are the regulations now in force governing the official branding 
of *' white-cure<r' herring in Scotland. They are presented ifi exfenso \ye<*&use of 
their thoroughness and the model they afford: 

Fishery B<>ard for Scotland. 

Regulations for examining barrels and half barrels intended to be filled, and branding and stencil- 
ing l>arrels and half barrels filled with cure<l white herrings, for the gnidance of fishery officers 
and the fish-c-uring trade. 

/. CajMtcity and tiuxle (tf nmHtrnction of iKirreU and half barrela filled or intended to be 

filled irUh cured white herrings. 

1 . Cajxtcitff: ( 1 ) Every barrel shall l)e capahle of containing 26^ gallons imperial measare. 
being e<inal to 3*2 gallons English wine measure. (2) Every half barrel shall be capable of con- 
taining 18^ gallons imi)erial mejisnre, lieing equal to 16 gallons English wine measure. 

2. Tight uemi: Every barrel and half barrel shall l)e perfectly tight. 

3. Staves and ends: 

(a) Thickness: The staves and ends of every barrel and half barrel shall, when completed, 
be not less than one-half part of an inch, and not more than three-fourths parts of an inch, in 
thickness throughout. 

(b) Breadth: (i) The staves of everj' Imrrel and half barrel shall not exceed 6 inches in breadth 
at the bulge, (ii) The head end of every l>arrel and half i>arrel must contain not less than three 
pieces and the lx)ttom end not less than two pieces. 

(c) Quality, etc. , of staves: The staves of every barrel and half liarrel shall be well seasoned 
and well fired, so as to bring them to a proper round. The staves shall not be cracked, broken, or 
patched, and there shall not be a double croze. The chime shall not be less than 1 inch in length. 

(d) Fitting of ends in crozes: The ends of everj' barrel and half barrel shall fit properly in 
the crozes, and shall not be turned inside out, nor bent outwards nor inwards so as to affect the 
sufficiency of the l>arrel or half barrel. 

4. Hooping: ( 1 ) Every barrel or half barrel shall l>e hooped in one of the three following ways, 
viz: (a) Entirely with wooden noops: (/>) entirely with iron hoops; or (c) partly with wooden 
hoops and partly with iron hoops. 

(a) Entirely with wo<Klen hoops: 

Ever>' Iwirrel c>r half l)arrel hoope<l entirely with wooden hoops shall l>e hooped in either 
of the two following ways, \\z: (i) Every barrel and half barrel shall l>e fuU-liound at the bottom 
end and have at least three go<^l hoops on the upper quarter, and everj* barrel shall have four good 
hooi)s and everj' half barrel three good hoops on the head end; the distance between the nearest 
hoops on opposite sides of the bulge of every barrel shall not exceed 1 1 inches after the hoops have 
l)een properly driven: the distance for half barrels shall l)e in like proportion. Or, (n) ever)* barrel 
and half barrel shall lie quarter-hoopetl, the barrels with four good hoops on each end and three 
good hoops on each <iuarter. and the half barrels with three g(xxl hoops on each end and three good 
hoops on each tjuarter. 

{b) Entirely with iron hoops: 

(1) Ever>' l)arrel hoope<l entirely with iron hoops shall lie hooi)ed in either of the two 
following ways, viz: (i) Every liarrel shall Ik? hooped with at least four hoojjs. one of those to be 
on each end of the Ijarrel and not to l>e less than 2 inches wide, of wire gage No. 16, and the other 
two to be on the quarters of the barrel and not less than IJ inches wide, of wire gage No. 17, the 
four hoops to 1r* placed at proper relative distances on the barrel. Or, (n) every barrel shall be 
hoope<l with six h(x)ps, one of these to l>e on each end of the l)arTel and not to be less than If inches 
wide, of wire gage No. 16, one to be on each of the quarters and not to be less than 1 inch wide, of 
wire gage No. 18, and one to be on each side of the bulge and not to be less than 1^ inches wide, 
of wire gage No. 17, the six hoops to be placecl at proper relative distances on the barrel. 

(2) Everj- half barrel hooped entirely with iron hoops shall be hooped with at least four 
hoops, one of these to be on each end of the half barrel and not to be less than 1\ inches wide, of 


wire gage No. 17, and the other two to be on the quarters of the half barrel and not less than li 
inches wide, of wire gage No. 18, the fonr hoops to be placed at proper relative distances on the 
half barrel. 

(c) Partly with icooden hoops mid partly with iron hoops: 

( 1 ) Every barrel hoope<l x)artly with wooden hoops and partly with iron hoops shall have 
either (a) the hoop of the head end alone, or (h) the hoops of both ends, made of iron at least 2 
inches wide, of wire gage No. 16. (a) If the hoop of the head end alone he of iron, the remaining 
jwrtion of the barrel shall be bound with wooden hoops in either of the two following ways, viz, 
the bottom end full bound, with at least three good hoops on the upper quarter, or (juarter hoopeil, 
with three good hoops on each quarter and four good hoops on the l)ottom end. {b) If the hooiw 
of both ends be of iron, each of the two quarters shall l>e bound with at least three good wooden 

(2) Every half barrel hooped partly with wooden hoops and partly with inm hoops shall 
have either (a) the hoop of the head end alone, or (6) the hoops of both ends, made of iron at least 
U inches wide, of wire gage No. 17. («) If the hoop of the head end alone be of iron, the remain- 
ing portion of the half barrel shall be Iwund with wooden hoops in either of the two following 
ways, ^^z, the bottom end full Iwund with at least three good hcwps on the upper quarter, or 
quarter hooped, with three good hoops on each quarter and three good hoops on the bottom end. 
{b) If the hoops of both ends be of iron, each of the two quarters shall be bound with at least three 
good wooden hoops. 

11. Marks which curers are required to put, or are prohibited from putting, on barrels 
and half barrels filled, or meant to be filled, with cured white herrings. 

(a) On the outside of the bottom of every barrel and half barrel, at the time when they are 
given by the curer to the packer to be packed with herrings, there shall Ik* legibly written or 
marked with red keel or black lea<l a description of the herrings to be packed, the date of their 
cure, and the number of the i)acker; and neither chalk nor any other substance shall be allowed 
to i)ass as a substitute for red keel or black lead, and no liarrel or half barrel unmarked as hert» 
prescribed shall be examined for branding. 

(?>) When any Imrrel or half barrel has been emptied of the herrings it contained, the old 
marks on the bottom shall be obliterated, and the barrel or half Imrrel, at the time it is given to a 
packer to be again packed with herrings, legibly marked anew, in red keel or black lead, with the 
description of herrings it is intended to pack therein, the date of the cure, and the numl)er of the 

(c) The curer's name and the name of the jwrt or place of cure shall be brande<l on the side of 
all liarrels or half barrels presented for the crown brand, and in addition the name of the district 
may be added thus: to Sandhaven may be addeil Fraserburgh, and to Boddam Peterhea<l, and 
a])ove these impressions there shall be legibly scrieve*! a description of the herrings contained in 
the barrels or half barrels, and the date of their cure — the month of cure to be expressed by the 
first letter thereof, except in the cases of January. April, May. and June, which shall lie designated 
by JA, AP, MA, and JE, respectively; the following being given as examples of scrieving: 12th July 
1895, La. Full, L 12 J/95; Full, F 12 J/95; Mat. Full, MF 12 J/95; Spent, S 12 J,'95; Mattie, E 12 J 95. 
On crown-branded barrels of herrings the year need not be branded, as that is given in the scriev- 
ing and also in the crown brand, which should l)e placed in close proximity to the curer 's name 
and the name of the port or place of cure. 

(d) No descriptive mark or marks shall be placed on the ends of barrels or half l>arrels of 
crown-branded herrings under penalty of removal of the crown brand without return of fees. 

///. Heading up of barrels and luilf barrels after filling them up with cured white hen'ings. 

After filling up, according as the barrel or half barrel has lieen ojiened at the head end or the 
bottom end, it shall be flagged round the head or bottom, made perfectly tight to contain the pickle, 
and pickled at the bunghole. The bunghole shall be ])ored within 1 J inches of the foremost lux)p 
of the left end; and both chime and quarter hoops of each end of every barrel or half barrel slitill 
be properly nailed. 


IV, Quality, method of aire, packing, etc., of white herrings necessary to secure the official brand, 

Qmility: The herrings shall be of good quality. 

Gutting, curing, and packing: They shall be gutted with a knife, and cured, and packed in 
barrels or half barrels within twenty-four hours after being caught. 

They shall be well cured and regularly ealted, and all fish broken or torn in the belly shall be 

They shall be carefully laid in barrels or half barrels, each tier l)eing completed with head 
herrings, and the herrings in each successive tier lieing arranged transversely to those in the next 
tier underneath, and drawn closely together, care being taken that the heads of the herrings are 
kept close t<^ the sides of the barrel or half barrel until it is completely filled. 

None of them shall l)e laid in bulk after l)eing cured in barrels or half barrels. 

They shall, if intended for the La. Full. Full, or Mat. Full brands, be pined in salt for not less 
than ten free days; and, if intended for the Spent or Mattie brands, they shall be similarly pined 
for not less than eight free days, these i^eriods to l>e exclusive of the day of cure and the day of 
filling uj) for branding; and this retiuirement shall apply to the herrings used in filling up as well 
as to those in the original packing. 

Filling up: 

{a) The surplus pickle shall be run well off through the bunghole, and the seastick herrings 
then left in the barrel or half barrel shall be pressed down by the cooper steadily and uniformly, 
by daunt or (use of daunt i)referable), thus testing the firmness of the original packing, 
and whether the surplus pickle has been suflSciently poured off or not. Pickle shall alone be used 
for the purjwse of washing herrings offered for the crown bniud. 

(6) The space left in the head end of the barrel or half barrel shall then be tightly packed with 
herrings carefully laid, regularly and lightly salted, the barrel or half barrel being firmly filled 
with herrings round the sides, as well as in the center. The herrings shall be pressed firmly to the 
sides of the barrel or half barrel with both hands, each tier being completed with head herrings, 
and the herrings in each successive tier being aminged transversely to those in the next tier under- 
neath, and the weight of the hands being pressed on each tier when finished, care being taken that 
the heads of the herrings in every tier are kept close to the sides of the barrel or half barrel until 
it is completely filled. 

(c) No herrings which have lost their original pickle shall be used in filling up. 

V. Conditions on which cured white herrings which hai'c lost their original pickle may 

secure the official brand. 

No herrings which have lost the original pickle shall be accorded the crown brand unless they 
have lieen repacke<l, washed in pickle, and presented separately for inspection, when if found 
worthy in every other respin^-t they shall, in addition to the crown brand, receive the '* Repack " iron 
across the St. Andrew's cross on the shoulders of the crown, so that it can not be removed without 
effacing the cro\\ni brand. If barrels or half barrels of repacked herrings, instead of l>eing offered 
separat^'ly, are found mixe<l up with any i)arcel of bung-i)acked herrings presented for the brand, 
the whole parcel shall lie rejected. 

VI. Iiea.viortment of rejected herrings for the official brand. 

When herrings once presented for branding have been rejected by an officer for bad quality, 
bad cure, bad gutting, or for being mixe<l with overday's fish (see penalty for presenting ov^erday's 
fish, etc. , on the back of request note, and at the end of these regulations) , they can not he reassorted 
and presented again for l^randing. 

Herrings rejecte<l for bad selection, or for too many undersized herrings for the standard of 
the iron applied for, may be reselecte<l and presented anew. Imt they must lie pickled with original 
pickle, when they may l>e crown-l)randed. if found otherwise satisfactory, with the *" Repack " iron 
added. The daunt must l)e used with all repacked herrings. 

Early-caught herrings: Herrings caught on the north and east coast of Scotland and on the 
coasts of Orkney and Shetland before r2th July shall not be crowTi-branded with the ** Mattie'' 
iron, while those caught on the coasts mentioned from 12th to 19th July, both days inclusive, shall 
have tlie long gut taken out before l)eing eligible for the ** Mattie " brand. 



T ! 







Winter-caught herrings: Winter herrings may, from the Ist November of the ^,^^€9 y 
Igt April of the succeeding year, be crown-branded, with the word '* Winter " branded r-^^i^^^ 
the St. Andrew's cross on the shonlders of the crown, so that it can not be removed withoiz^s. e/ 
the crown brand. 

VIL Examination of barrels and half barrels in respect of their capacity and mode ofconstn 

(a) Barrels and half barrels intended to be filled with cnred-white herrings: Office] 
examine at least fonr in every hundred barrels or half barrels intended to be filled with he 
the capacity of one being tested (if necessary) by liquid measure, and the capacity of the i^, 
ing three by diagonal rod 23 inches long for barrels and 18^ inches long for half barrels, me^s 
from the croze of the bottom end to the croze of the head end; the examination to l)e mad< 
time or times suitable for the officers themselves. 

(6) Barrels and half barrels filled with cured white herrings: Officers shall examine all barr^ 
and half barrels filled with herrings, and (if necessary) shall empty the herrings out of at le^"^ 
one barrel or half barrel in every hundred and test its capacity by liquid measure and test t^^ 
capacity of at least three others by callipers. 

VIII, Examination for branding and stenciling barrels and half barrels in respect of tJie 
quality, method of cure , packing, etc., of the white herrings they contain. 

(a) The barrels or half barrels presented for branding shall be laid out so that the bottom endR 
come at once under the eye of the branding officers. 

{b) The curer, or his authorized manager at the place of cure, having delivered to the officer the 
prox>er account of cure of the herrings presented for branding, and a request note containing the 
number of barrels and half barrels to be presented, the officer shall see, first, that the number of 
barrels and half barrels is correctly stated in the request note; second, that the request note is 
signed by the curer or his authorized manager (as the case may be) ; and third, that the branding 
conditions attached thereto are likewise signed by the curer or his authorized manager. 

N. B. It shall be understood that no manager can be recognized as an authorized manager 
except under authority obtained from the board upon application previously made by the curer. 

(c) Brand fees (at the rate of Ad. per barrel and 2d. per half barrel) corresi)onding to the 
correct number of barrels and half barrels in the request note shall be deposited with the officer 
before branding, subject to the condition that if the x)arcel be not branded the amount of brand 
fees so deposited shall be returned to the curer; or. if only a portion of the parcel be rejected, the 
brand fees corresix)nding thereto shall be returned. 

(d) The curer *s declaration shall then be taken and signed by him. 

(e) The minimum number of barrels to be examined per hundred shall be seven. Two barrels 
X>er hundred or smaller parcel shall be examined down through the original, the remaining five 
down to the lower quarter hoop of either end. 

Officers are not restricted to this scale, but if need be shall ox)en as many more barrels or half 
barrels as they may deem requisite to satisfy them that the herrings are fit for branding, for 
which they will be held resix)nsible to the board; but they shall understand that in no case what- 
ever shall fewer barrels or half barrels than what is prescribed in the above scale be ox)ened for 
examination previous to branding. 

if) The barrels or half barrels selected for examination shall, as a general rule, be opened at 
the bottom and head end alternately — that is to say, No. 1 shall l>e opened at the head end. No. 2 
shall be opened at the bottom end, and so on until the whole examination is concluded. The 
herrings in all barrels or half barrels ojwned shall be searched down to the lower quarter hoop 
of either end, two barrels per hundred or under as in note (e) , and as uiuch farther as may bt* 
deemed necessary. 

But where an officer, from any cause, sees reason to examine a larger proportion of barrels or 
half barrels at the one end than at the other, he shall be at liberty to substitute the examination 
of such larger proportion for the above alternate examination, only observing that not less than 
the full proportions per hundred or smaller parcel which are laid down for the minimum scale of 
examination shall be examined in all, and that as many more shall be examined as he may see fit. 

(gf) In examining a parcel the work of different packers shall be selected, as well as herrings 
of different dates of cure. 



(h) All objectionable herrings shall be removed from the barrels examined before afiizing 
the crown brand. 

(i) When, on the first examination of herrings for branding, they are found of bad quality, 
badly cured, or badly glutted, refusal of the brand shall be final and absolute. When, however, 
this refusal has been entirely owing to the liarrels or half barrels being too slackly x>acked with 
herrings, or the filling up badly selected, the case shall l)e treated exceptionally and shall be reme- 
died by filling up or reselection only of the filling up, and the herrings may thereafter be branded, 
if in accordanct^ with the following conditions: 

( I) Tliey shall he ])resented for renewed inspection only to the officer who previously rejected 
them, who shall satisfy himself by full examination that they are, apart from slack 
X)acking or bad selecticm of the filling up, worthy of the })rand. 
(ii) The filling up shall have been properly comi)leted; but. failing this, the herrings shall be 

finally rejected and no further examination permitted, 
(ill) The officer shall retain the fees until the herrings are branded or finally rejected; in the 

latter case returning the fees, 
(iv) He shall state uxx)n the request note the particulars of the first refusal; and if the her- 
rings be afterwards branded, the date of branding. 
When, however, a parcel is small, and \\\)OJi the first in8i)ection the deficiency in filling up is 
seen to l)e so very trifling that it can be supplied at once in the presence of the officer without 
difficulty or detention, the alK)ve conditions shall not api)ly, but the filling up may 1)8 done ujion 
the spot and the branding proceeded with immediately afterwards, the officer being careful to 
satisfy himself jirevious to the ])rauding that the herrings are in all other rt»sx)ects entitleil to 
be branded. 

(k) The officers shall see that the barrels ox)ened are filled up and headed with proper care. 
(Z) The officers shall put a double crown cm the bilge of the ])arrel examined and toward the 
end examined. 

(in) Oversalting shall be determined by the measure known as the cog; and the quantity of 
salt left in any barrel emptied of fish must not exceed this measure. 

IX, Branding and stenciling barrels and lutlf barrels in respect of the quality y method 

of cure, packing, etc., of the white herrings tliey contain. 

Every barrel or half barrel containing white herrings presented to one of the officers for exam- 
ination shall, if the cai^acity and mode of construction of the barrel or half barrel, and the quality, 
cure, selection, packing, etc., of the herrings are, in liis opinion, such as to satisfy the requirements 
of these regulations, (1) have branded in his presence, by means of a hot iron, on the bilge, in close 
pn>xiniity to the curer's name and the name of the port or place of cure, a (trown surrounding the 
word *• Scotland,'* a description of the herrings, viz: La. Fnll, Full, Mat. Full, Spent, or Mattie 
(as the case may be), the initial letters of the examining officer's name and the year; and (2) have 
stenciled in his presence, cm the head end, a crown surrounding the same word, description, and 
letters as those branded on the bilge, with the words ** Fishery Board, Crown Brand,' stenciled 

X, Requirements of the different Itramls, 

In addition to what are contained in the fon^going regulfitions, the requirements in resi)ect of 
the different brands shall \ye as follows: 

Crown '*La. Full" Brand. 

Barrels or half barrels of herrings f(>r this brand shall ctontain large full herrings of not less 
than 1 1| inclitjs in extreme length, as measured by the fishery officer's gage, made for the purpose. 
Rejections under this brand sliall be: 

(1) On original x)acking for spent, torn, broken herrings, or herrings of bad or indifferent 
quality if more than fifteen; or. on filling up, if more than six. 

(2) On original packing if the undersized amount to more than fifteen; or, on filling up, to 
more than six. And the parcel shall also be rejected if it should appear that the larger herrings 
suitable for this brand have been previously taken out. 


Crown **Fuir' Brand. 

Barrels or half barrels of herrings for this brand shall contain fnll herrings of not less than 
lOi^ inches in extreme length, as measnred by the fishery officer^s gage, made for the purpose. 
Rejections nnder this brand shall be: 

(1) On original x>acking for spent, torn, or broken herrings, or herrings of bad or indifferent 
quality, if more than eighteen; or, on filling up, if more than nine. 

(2) On original packing if the undersized amount to more than eighteen; or, on filling up, to 
more than nine. And the parcel shall also be rejected if it should appear that the larger herrings, 
suitable for this brand, have been previously taken out. 

Crown *'Mat. Full" Brand. 

Barrels or half barrels of herrings for this brand shall contain full herrings well develope<l — 
the roe or milt being clearly seen at ne<;k or throat without pressure — of not less than 9| inches in 
extreme length, as measured by the fishery officer's gage, made for the purpose. Rejections 
under this brand shall be: 

(1) On original x>acking for spent, torn, or broken herrings, or herrings of bad or indifferent 
quality, if more than twenty-one; or, on filling up, if more than nine. 

(3) On original x)acking if the undersized amount to more than twenty-one; or, on filling up, to 
more than nine. And the parcel shall also be rejected if it should appear that the larger herrings 
suitable for this brand have been previously taken out. 

Crown ** Spent '' Brand. 

Barrels or half barrels of herrings for this brand shall contain spent herrings of not less than 
lOi inches in extreme length, as measured by the fishery officer's gage, made for the puri)08e. 
Rejections under this brand shall be: 

(1) On origfinal packing for torn or broken herrings, or herrings of bad or indifferent quality, 
if more than eighteen; or, on filling up, if more than nine. 

(2) On original packing if undersized amount to more than eighteen; or, on filling up, to more 
than nine. And the parcel shall also be rejected if the larger herrings have been previously 
taken out. 

Crown '* Mattie *' Brand. 

Barrels or half barrels of herrings for this brand shall contain herrings not eligible for any of 
the foregoing brands, and of not less than 9 inches in extreme length, as measured by the fishery 
officer's gage, made for the purpose, but shall not contain headless herrings. Rejections under 
this brand shall be: 

(1) On original packing for torn or broken herrings, or herrings of bad or indifferent quality, 
if more than thirty; or, on filling up, if more than twelve. 

(2) On original packing if undersized amount to more than thirty: or, on filling up, to more 
than twelve. 

''Repack ''Brand. 
For exx)ortation out of Euroi)e: 

1. The herrings for this brand shall have been pined in salt for not less than ten days, exclusive 
of the day of catch and the day of beginning to repack for branding. 

2. They shall be emptied out of each barrel or half barrel in which they were originally cured, 
and they shall be washed clean. 

3. They shall have the crown gut, if adhering to them, removed. 

4. They shall be repacked into the barrels or half barrels from which they were emptied and 
into as many additional Imrrels or half barrels as may be necessary. 

5. They shall be salted sufficiently, and be pickled with strong pickle made of clean salt. 

6. Every barrel or half -barrel shall be full bound at the head end as well as at the bottom end, 
and shall have at each end an iron hoop of 1 inch in breadth. 



** Lozenge " Brand. 

1. This brand shall be applied to herrings previoa««ly branded which have been repacked in the 
manner retjuired for the " repack " brand, and the lozenge shall be stamped immediately under and 
close to the crown brand already npon the barrels or half barrels. 

2. Upon the additional barrels or half barrels derived from the repacking the *' repack *' brand 
shall be affixed, subjoining thereto the lozenge brand as above. 

Note. — By the strict letter of the act the cnrer or proprietor of the herrings ought to give 

twenty-four hours* notice, in writing, of his intention to repack for this brand, but, of 

course, where the officer can accomplish his examination of the herrings sooner he 

should endeavor to do so and accommodate the curer as far as he can. 

Under any crown brand, if the officer is satisfit^l with the c^ire, quality, etc.. of the herrings, 

but considers them generally too flatly packed, in addition to the crown brand he shall cause the 

lozenge brand to be affixed to cover the St. Andrews cross on the top of the crown. 


WB 95 


J.M 95 






J.L 95 



Representations of Brands used in Scotland in markinu Barrels of Glared HERRiN(i. 

A)j<jut onc-balf actual tazv. 



Vlaardingen, situated on the Maas, a few miles l)elow Rotterdam, is the center 
of the Dutch herring trade. There the herring boats fit out, there they land their 
catch, and there are the houses in which the fish ai"e pn^pared for shipment. 

The herring fishery is conducted by steam and sail vessels, which use tanned 
cotton gill nets 3G0 meshes deep and 720 meshes long, the size of mesh l>eing 2-ineh 
stretch. From 80 to 150 nets are carried by each vessel, this outfit usually costing 
from 5,000 to 7,000 guilders. The nets are set about 6 feet below the surface, being 
held in position by leads and by corks (5 inches long, 3 inches wide, and 2 inches 
thick) at intervals of a foot, the cork line being attached to the top of the net by 
numerous gangings 8 inches long. At times, when the herring are close inshore, 
some surface fishing is done. 

The dressing and salting of the fish immediately after the nets are hauled on 
the vessels are considered of great importance by the Dutch herring fishermen, and 
no doubt contribute largely to the quality of the cured fisli. 

The Dutch method of cleaning herring is similar to the Scotch. Provide<l with 
a short knife, attached to the fourth and fifth fingers of the right hand by a string 
tied to the handle, the fishermen take the herring in the left hand, with the l)elly 
up and the head forward, and thrust the knife crosswise directly through the gill 
cavities, entering the left side and emerging from the right. The edge of the knife 
being turned upward or outward, the knife is pulled directly through the tissues, 
cutting and tearing away the gills, branchiostegals, heart, oesophagus, stomach, and 
often a part of the intestines; the pectoral fins, with the skin and muscle at their 
base, also ciome away with the same movement. There is apparently little effort, 
made to remove anything except the gills and pectorals, the other organs coming 
away incidentally. The men become very expert in cutting, and some of them can 
handle 1,200 fish per hour (20 per minute). 

The removal of the gills and heart results in opening the large blooil vessels, and 
free bleeding ensues; this leaves the flesh pale or white, in contrast with the dark 
reddish cx)lor of the Scotch herring, in which the blood is allowed to clot. It some- 
times happens in the Dutch fishery that when there is a large catch the blood has 
clotted in the last fish handled. The chief and only genuine l>enefits of cutting as 
now practiced are the bleeding and the opening of the alKlominal cavity to the brine. 

Some herring examined by me as brought in by the fishing vessels at Vlaardingen 
contained pyloric cteca and part of the intestiiies, as well as the liver and reproductive 
organs. The intestine, with or without the cjeca, often hangs outside th(» wound 
made with the dressing knife; it is called the **zeele" (soul), and is frequently eaten 
by the packers of salt fish, being regarded a choice morsel. 

The packing of herring is done while the vessels are still at sea. The fish are 
first rolled in salt and then carefully packed in straight rows, with backs down. 
The fish in a given layer are at right angles to those of adjoining layii^i's. One barrel 
of St. Ubes salt is required to pack four barrels of herring at sea, the salt iHMug 
disposed between the layers of fish. The barrels are headed up and stored in the 
hold until the fishing trip is ended or all the barrels filled. On reaching port the 
catch is unloaded and sold at auction. 


The buyer almost always repacks the fish in oiiler to sort them by size and 
grades of quality, no sorting being attempted on the vessels. Sometimes purchasers 
or agents prefer the sea-packed, unsorted fish, but as a rule the dealers or jobbers 
wish to know how many fish are in a barrel and what their quality is. Some 
shrinkage ensues; this is usually made up with fish of the same lot before the sale, 
the refilling being done either on shore or on the vessel. 

After coming into the hands of the packer the herring are emptied into large vats 
or tanks, from which they are repacked according to the prevailing practice. The 
original brine (called *' blood brine" or ** blood pickle") is considered much better 
than any newly-made brine, and is always saved and poured back on the fish after 
repacking. The fish are placed in the barrels in the same manner as at sea, and 
fresh salt is added in the proportion of 1 barrel to 8 barrels of fish. The shrinkage 
in repacking is about 8 per cent — that is, 100 barrels of sea-packed hernng will 
make about 92 barrels of fish ready for market. 

The Dutch herring barrels conta>in about 125 kilograms of fish, and most of the 
catch is marketed in such barrels. Smaller receptacles — J, i, i, ^V> ^"^ A barrels— r- 
are also used, but are not nearly so much in demand as they were a few years 
ago. The barrel staves are oak and are imported from New York, Baltimore, and 
Newport News in the form of rough pieces, which are cut into proper shape at 
Vlaardingen, where several thousand coopers are regularly employed. The hoops 
are made from willow trees grown on the dykes. White, clean barrels are required 
for the American trade; dark, dirty barrels are ac<?epted by the continental countries. 
The Scotch herring barrel is regarded as a very good, strong barrel, and is imported 
by the Dutch packers. The preferred arrangement of the hoops is to have four 
between the bung and each end as well as four at each end, so that when the barrel 
is rolled its weight rests on the hoops. The bung is large and central. Some barrels 
have a single iron hoop at the top. 

In Holland there is no official I'egulatioii of packing or branding, but the packers 
have a standard which is generally observed, as it is to their interest to have the 
fish proi)erly packed and labeled. The diffei-ent grades of herring recognized are 
similar to the Scotch, and are based on the spawning condition of the fish. The ripe 
or full fish are branded *'vol" (=full); the matties(maatjis) are branded "m"; the 
spent herring (ijlen) are branded " ij"or "ule." Of each of these there are several 
qualities designated No. 1, No. 2, etc., and there are several other grades. The 
barrels are usually marked with a stencil. 

The Dutch herring trade is not restricted to the fish caught and packed by the 
Dut<;h fishermen. Considerable quantities of salt Scotch herring are received at 
Ylaardingen, which, after repacking, are sold as Dutch herring. Furthermore, the 
Dutch sell some of their own herrings in Scotch barrels in the continental countries, 
where the Scotch pack is well known. 



The experience of the European herring-paekera has result-ed in a prepared 
product, which meets witli ready sale throughout tlie world at Ix^tter i)rices than are 
received for otlier cured herring. If American herring-curers wish to 8Ui)ply the 
home markets and to establish a profitable trade with other countries, tht»y must 
take cognizance of the demands of those markets and make their fish conform thei^eto. 
That there is an opportunity for a largo increase in l^oth the domestic and foreign 
trade in American herring there can be no doubt; and the following suggestions are 
made to this end : 

1. While the demand for fresh herring for bait, for smoking, and for canning 
tnkes a large part of the catch on certain parts of the New England coast, there are 
localities where the salting of herring could be made vii^ry profitable. Even in the 
canning and smoking districts it may prove more remunerative to the weir fishermen 
to salt their large-sized herring. It seems jirobable that the excellent herring of the 
Pacific coast can l)e salted to great advantage and ought to find a ready market. 

2. Care must be exercised in all steps of the curing and i)acking processes. 
Only plump fish in the best condition should be salt-ed, and only sound fish should l)e 
packed. Herring of different grades should not be packed in the same barrel. 

3. Different standard qualities of salt herring should l)e recognized and con- 
scientiously adhered to. The organization of the United States Government would 
probably not warrant the Federal authorities in exercising jurisdiction in the matter 
of inspecting and branding fish. While this jurisdiction could doubtless be acciuirinl 
(as has been done in the case of meats intended for exi)ort), it can not l)e reganhKl 
as essential. Each State is comiietent to superintend the inspection and branding 
of its own fish, to adopt special regulations and brand marks, and take such other 
measures as will tend to promote the salt-herring trade. This syst-ematic branding 
under State authority is regarded as one of the most essential factors in the develop- 
ment of the salt-herring trade. 

4. The establishment of a large tra<le with southern Europe, the Philippines, 
Australia, and elsewhere in salted river herrings or alewives is entiivly feasible. 
These fish, which are in excellent condition when they ascend the Eastern rivei-s in 
untold millions each season, should, if properly prepared, sell almost hh well as the 
sea herring. An especially good opportunity for i)romoting the ale wife fishery 
appears to exist in the Middle and South Atlantic States, where the cat<5h is only 
imperfectly utilized and where labor is abundant and cheap. The river herrings 
might be prepared as white-cured herring and also as red herrings. 


Adjunct fYo/essor in Zoology, Columbia University. 

Eun)peHii oyster-culture, esi)ecially as practiced in France and Holland, is gen- 
erally regj'rded as the most refined development of this ancient art; and from our 
knowledgeof this — if we admit that those cultural methods are the most i>erfeet which 
produce the greatest number of oysters in a given area — we can reasonably conclude 
that some of the devices at least of the European culturist will ultimately come to 
be adopted in our own system. In view, accordingly, of the prospective value of 
foreign methods in the development of American oyster-culture, the United States 
Commission of Fish and Fisheries has already published (in its bulletins^ for 1890 
and 1801) reports \\\^n\ the practical workings of the best forms of European oyster 
parks. From the character of the methods there in use, we can, I believe, conclude 
positively tint similar establishments could be operated successfully at suitable 
points — e. g., in Chesiipeake Bay or Long Island waters, as soon, that is, as the 
deman<l for oysters will warrant the use of what will prove at first a moi-e expensive 

While these Euroi>ean methods are applicable on our Atlantic coast, it still 
remains to be determined whether they include the l)est that could lx.» employed 
along the Pacific, should artificial oyster-culture l>e here attempted. For in these 
watei*s different conditions have prinluced oysters which differ widely from those of 
the Xorth Atlantic. The Pacific culturist mav therefore feel a more livelv interest 
in the oysters of Japan, for not merely are they closer akin to his own, in structure 
and in habitat, and therefore more readily acclimatable, but they are larger, better 
shaped, and certainly of greater value, commercially siJ4*aking, than the local product, 
Ostrea caUfoniira. Moreover, the Jai)anese oysters have long been cultivated, and 
with great success. Indeed, by some ex|>erts the Japanese methods have l>een 
commende<l as the simplest and most practicable of all forms of "artificial" oyster 
culture, an<l thus of j)Ossible interest in somewhat broader lines. 

Unfortunately there is no litemture accessible dealing in detail with the culture 
or living conditions of this western Pacific oyster, and it is w ith the aim of filling 
this gap that the present report was prepared. Its material was collected by the 
writer during a stay in Japan in rJOO-1901. He there acted under special instructions 
from Commissioner George M. Bowers, and in aid of his inquiries was designate<l as 
a biologist of the United States Commission of Fish and Fisheries. 

« R«.»port on the Prej^ent Methods of Oyster-culture in France. Bulletin U. S. Commission of Fish and Fiaherie«, 
taeu, ijages 3«:i-:^, plat«-s Lxvui-Lxxvfu. 

Report on the Eunjpeau Meth<Kls of Oyster-culture. Op. cit. for 1«W1. i>age8 357-4<Jt>, plates LXXV-LXXXVIII. 
(Italy, Spain, Portugal, Germany, Holland, Belgium, and England. > 

F.C.B.19a»— 2 


For the rest, Japanese oyster-culture proved to be wortliy of careful study, not 
only for its merits, but because of the suggestions it affords for cultural experiments. 
One may frankly doubt whether it can at once be employed profitably — for example, 
at many points on the Oregon coast — in view of the expense for labor which it 
entails, but I believe that there is a reasonable chance that it could be made profitable 
if employed in a favorable locality. In any event, so far as the Pacific coast is 
concerned, the Japanese methods are the most practicable, and cxi)erinients with 
them could be made readily and at little expense, and would soon demonstnite 
whether artificial oyster-culture can here be employed commercially. 

The Japanese industry is largely seated along the north shore of the Inland Sea 
near Hiroshima, in the gulf-like Sea of Aki, famous for it« oysters. From what 
period, indeed, this oyster-culture has been wirried on is not known accurately, but 
from its present condition it is evidently the pro<iuct of centuries. As early as 1708 
records show that conc^essions at Osaka were granted to an oyster company or to 
oyster companies of Aki for storage of their output pending the final marketing." 

Regarding the origin of the oyster-culture in this region I may here (juote a 
paragraph from a tract on fishery matters published by Iliroshima-Ken. 

In ancient times certain shellfish, Tapes , were gathered in great nniubers on the flat^^ of Aki; 
and while awaiting their shipment to market the fisher people cAiue to keep them in shallow- water 
inclosnres, the fences of which they formed of bamboo stalks. The discovery was then uia<le that 
the brushy fences became incmsted with young oysters, and thus it soon became evident that 
under certain conditions and at certain places it would be more profitable to plant luimlxM) and to 
cultivate oysters than to continue the Tapes industry. This was the first instance, it is said, that 
bamboo collectors, or *' shibi," were employed in oyster-cmlture. 

The first detailed report upon the oyster industry of Aki was prepare<l for the 
Japanese government by Prof. Kakichi Mitsukuri, the head of the department of 
zoology of the Imperial University of Tokyo. It was published in lSt)4 (Tokyo) by 
the department of agriculture and commerce, a royal octavo of al)out 50 pages, 
containing many figures and several plates.* Unfortunately for the foi'eign reader 
it has not yet been translated into a European language. My own knowledge of it 
is due to my friend (who has also kindly drawn for me the text figures here repro- 
duced), Mr. Naohide Yatsu, Rigakushi, a pupil of Dr. Mitsukuri. I have availed 
myself freely of it« substance, and if there is anything of value in the present paper 
it owes its merit to my Japanese colleague. lie has, moreover, given me generous 
assist>ance in connection with my visit to Aki, advising me as to ways and means 
and providing me with personal and official letters to the local au^Jiorities. 

In Hiroshima I was courteously received by his honor Governor Senshi p]gi, of 
the prefecture of Aki, and to him and to his staff I am indebted for favors ext<Mided 
me in many ways. To Mr. Shinobu Suda, official engineer of tlie oyster properties, 
and to Mr. Masugi Shiraishi, a government expert in oyster matters, I am under 
especial obligation for information regarding details in cultural methods; and 
finally to Mr. Y. S. Murai, for many |)ersonal courtesies. 

o A probable trace of this early privilege is seen at the i>resent day, for tho iHM>ple of Aki are Krantecl the best 
places in the river mouths of Osaka. 

t* All the present text figures, except flg. 27, are reproduced with unimiM>rtant chanKi^ from tlie Jai)ani>Ho t eport. 
Plates 4 and 5 are from photographs taken by an artisi in Hiroshima, and are cui>iod by Dr. Mitsukuri; Plates 3 and 



Three kinds of Japanese oysters are to be considered in this connection. First, 
a small one, probably a dwarfed salt-water variety of Ostrea cucuUata Born., is 
abundant along the southern and eastern coasts of Japan — by far the most abun- 
dant of its kind. This is a bay oyster, wcurring in shallow water of specific 
gravity of about l.o^o to l.o^G, where it forms an almost barnacle-like incrustation 
upon the tidal rocks. It is collectiMl in great numbers for local consumption; fisher 
people oj)en them on the sjyot, not detaching the shell from the rock, and market 
them by bulk. In actual size this oyster is rarely larger than one's finger nail; but 
its flavor makes amends for its size. It is pluml>eous in color. The shell measures 
rarely more than 2 inches in length; it is deeply crenulate, Gryphaea-like, at the 
margin. In size and flavor it suggests very closely the California oyster. 

The second form, Ostrea cucnlhda (cf. pi. 0, fig. D), is the important one from the 
culturists' standpoint. Although not large, it averages the size of a "Blue Point," or 
of an English '' native.'' The oyster itself is cream -colored, it« shell delicately nacre- 
ous, well shaped, thin, deep, and with a series of imbricating, horn-like outgrowths, 
which suggest the shell of the Euroj)ean oyster, (). edulis. This species occurs 
abundantly throughout the Inland Sea, in the snuill bays along the northeast coast 
of the main island an<l at certain points in the Hokkaido (Yezo). It thrives best 
in the bays well tampered by fresh water, of si)ecilic gravity of 1.017 to 1.023. Its 
young are more abundant in the shallow and fresher water. The best that are mar- 
keted grow at a depth of a fathom or two l>elow low-water mark; it is practically 
absent in water deeper than 8 fathoms. It is this si)ecies which will be considered 
through the remainder of the present report in connection with cultural methods. 

The thii-d form, Ostrea gigas Thunb., is of large size, specimens weighing with 
shell 4 or 5 pounds being not infi-equent. It rarely occurs in water less than 2 
fathoms deep and is most abundant in alK>ut 10 fathoms. The si>ecimens which I 
examined wei-e taken by divei-s in water of about 35 feet. It is a typical sea oyster, 
occurring in water of specific gravity of about 1.02G. As far as I have l>een able to 
ascertain, its value is purely local, no region producing sufficient numbei's to warrant 
a definite fishery. A large bank of oysters occui-s in the Hokkaido, off the north- 
east coast, not far from the town of Akkeshi. The oysters here are said to be of 
extraordinary size, but during my visit to the Hokkaido I was not able to ascertain 
whether they represent this third species or whether they are large examples of O. 
cucuUata. The latter species certainly occurs in the neighlwrhood. 

The oyster-producing region of Japan is jtar e.rcellence the Inland Sea, and it is 
here that the culturists have carried on their industry with greatest success. This 
body of water can indeed be looked upon as one of the most important natural pre- 
serves of fish and shellfish in the world. It can be compared to a deep marine lake, 
but it is sufficiently open to the sea to insure favorable conditions of density and of 
renewal of its waters, while its occupants are free from the dangers of an open gulf. 
From the oyster-cult urists' standpoint the Inland Sea is remarkable in that its con- 
nection with the ocean is establishe<l both at its en<ls and near its middle point. 
Thus at the extreme east it oj>ens to the ocean through the Straits of Naruto, as 
well as at th(» mouth of the Izuminada. At the west, 240 miles away, it opens again, 
this time to the Japan Sea, through the Straits of Shimonoseki, and to the south 



again to tbt^ I'licitic through Biingo Clianticl. In tliiu tniddlu region the lai^e isliiud 
Shikokii approa<!heH closwly the luainlaiKl, anil the Inlantl Soti is hroln-ii up by n 
maze of islandK extending from ShodosliiniH on the <-aHt t.o Iwal on the went, a 
stretch of 130 miles; and it iis Iktc that th<' iiioNt favoralili; conditions exist for tlic 

PlO. 1.— Mftp of the rogion of Jupanoso oyster-culture on the nurth iJiorii of the- lulaiid Si 

(Sea of Aki). Oyatcr parlu art- in<UoAt<Hl lu lilark ar<nn. Partli^ular r»ri'n'n<'>^ In lli<' pn-wtit r-'l'-rl li> 
Kaida Bay. Kusatsu. and Nihojlma. Tbi.' diiitan<x> from Kalda to tho Inland ItKiikiwIiiiiiii i^ 13 iiiil<.i<. 

growth of shelllisli. Everywhere are bays anil liarbors, and tho density of tlie 
shallow wat«r is favorably teiuiK^ivd by the iiieoining stn-anis of Aki, l!iiigo, ItiK'hu 
and Bizeu on tlie north and lyo and Hauuki on tlie Houth. Add to these advant^tgeH 
that there is a favorable tide fall of from 10 to L5 feet and an abnndauuo of sandy 


and gravelly bottom, thus cnabliug the ciiltiirist t« operate his 8iibmei^re<l farms 
couveiiiently. Throughout this entire region oyster-cult lire in carried on more or 
less generally, but the most important seats of the industry aif at Okayama in the 
east and near Hiroshima (prefecture of Aki) in the west. In the former locality a 
small nearly- in Hosed bay, which suggests that of Arcaclioii in France, proves very 
pnKluctive and supplies no little part of the market of Kol)e and Osaka. Here 
also are canning factories. Near Hiroshima, however, the industrj- is conducted on 
a stmiewhat larger scale, although on the same genei'al lines. 

It is in the latter n'gion, iis ah-eady noU-d, that Professor Mitsukuri secured the 
material for his n-port itynm Japanese oy.ster-cullure. 


In addition to its natural advanUij^es ihe region of Aki is esi)ecially favorable 
for oyster-culture, since close by is Hinishinia. a city of nearly lOd,^^', to furnish a 
ready market for its i>ro<luet and to provide the necessary capital and lal>or for the 
growth of Ihe industry. Tlic culture is oarnc<l on in a gidf-likc area, known as the 
Sea of Aki, a <lozcn miles in wiillh, whose mouth is protected from southern storms 
by the islands of Itsukusliinm (Miyajima), Nomijima, and Etajima (fig. 1). The 
most favorable points for the cultural work are along the northern c<iast on either 
side of the Otagawa, which flows thn>ugh Ihc city ami tempers the .salinity of the 
neighlKiring water. On the east of Hiroshima are the establishments of Nihojima, 
and somewhat further up the bay, Kaida; to the west, the grounds of and near 
Kusstsu. The entire est^-nt of these most favorable points is about 10 miles. This 



cultural area, it may be not«<1, corro»pond» closely to that of Tarente in the Houth 
of Italy, of Arcachoii in France, and of the best part of Tiong Island Sound. 

It is an interesting fact that the culturiBts in Aki have at certain points developed 
independently branches of the indiiatrj' which are strikingly similar to thOHC employed, 
for example, in France. We thus find iliat n clear <listinctinn is inatle Iwtween Ihe 
regions in which young oysters — " spat " — can Ik" obtained and those having the l)est 

Fio. a.— Hand pink owd for 
maklnK ra-ieU in gi»Telly 
bottom fur Inwrtlnii "f 

.conditions for diffeivnt stages of gi-owih; also regions in which llie final Uiuches are 
given in pn>paring the oyster for nuirket. It is eonveiiieiit, tlierefon', lo dewrilje 
the Japanese cultural methods from the standjH^int of loeality. Thus nt Niliojima. 
where the waller is fn'sheiied by the entrance of the OtjiKawa, the "pnMhiction" of 
young oysters is an especial feature of the iiidustiy. At Kaida Itay then- is a region 
favorable for a combination of production and growth (elevage), and at Kusa1su,and 
further along in the direction of Miyajtnia, ai-r' Ihe Ix'st conditions for elevage. 



In the description of the methods employed in these three localities, it will be best 
to consider them in the order from the simple8t and the most complex, (1) Kaida, 
(3) Kusatsn, (3) Xihejima, 
for at Kaida Baj' all stages 
of the induslry are repre- 
sented in the same oyster 
park; at Kusatsu tlie 
methods l>ecome more 
complicated, and finally, 
in the region of Xihojima, 
specialization in the cul- 
tural devices has reached 
a point surpa.'tsed in hut 
few European localities. 


In this well-protecteti 
region (cf. the accompa- 
nying map of the sea cf 
Aki, fig. 1) there is a large 
area of shallow water, and 
at h>w tide great flats are 
exposed. Here ithasl>een 
ascertained that the con- 
ditions of water density 
are favorable rather for 
the growth of young oys- 
ters than for the prmliic- 
tion of "spat," but at cer- 
tain i>oints production is 
carrie<l on with niarke<l 
success. Thegreatest dis- 
a<lvantage of the region is 
the lack of space in which 
the oysters can l>e kept 
covere<l by water during 
all times 'of the month. 
Probably it is for this i-ea- 
son that the gmwthof the fio. b.-a 
oysterc-onies1ol)eehecked, /[^^^ 
since it is well known that "".n in 
they rarely increase in size v^'^u-aT 
after the end of the second 
year. At this age, then, they are marketed, their small size distinguishing them 
fnim the oyst^ers of Aki cultivated in other localities. The hay of Kaida is, however, 
so fertile in its class of production that it takes a high place among Japanese oyst«r- 

. :.— Shibi of different rows, new and old. in 
madary hi^^, tUiDwins how they are tm- 

anted to (fire matual support. 



grounds and its concessions are keenly sought. At low tide it bristles with closely 
set oyster farms and from a distance reminds one, save in color, of a region of 
European vineyards. Each farm is a simple inclosure formed by '* shibi " or bamboo 
stalks, with or without interlacing branches. (Cf. figs. 4 to 0.) Bamboo in this, as 
in many other arts and trades of the Japanese, possesses many advantages. It is 
durable even in sait water (good material lasting three years or thereal)outs) ; it is 



/ --4-- 




^~ 1 





-- i 


— _._...._ 

* __ _ 1 



— • — . 


1 g J 

Pio. 8.— Typl*^! oyBter farm, Kaida Bay. Fio. 9.— Diagram of woll-<lev*'loi)ed oyster farm. 

The blu4;k lineH In flgH. 8 and 9 represent newly arranK«*<l bamlxxi <*oll»»<!torH, tlu? dc»tt<Hl lin«»H 
the oollector« of the Hcnumd year. Direction of current in indicated }>y arrows. 

light and strong, gives an interlacing series of branches and heaves wliich in texture 
serve admirably for the attachment of spat, and which give, moreover, a gn»at <»xt^nt 
of attachable surface. In addition the bamboo stalks can 1k» n»a<lily put in phwM* and 
removed; they are easy to obtain in any locality, and their cheapness is not one of 
their smallest virtues. 

In the present farms shibi are plant^nl in position every springata time which the 


M^^^^B^^ZSui^BBB^^bii 1^ 'rSHE^^H 





»vi'i' Vp-? * ■■'■i*'*^v^''^»i' ■ 

1 s 





Be ' ^y^^JBKiOBB'iWr^^^ntfvliilS 4 ^^^^^T" ^ /r ^B^^H 


B^ * THSaBpyy " ^' y ^ft H^ ,iX^ p1 .i''^j|^^H 


^9^ ' '' 'f ~- ' '*^^' ' ' •' •'7'^_f ^^B^^^^^^^^Hm^I 

H i 

^k " c*>ift ^ V'' ^ihI^^^I^^^^^^B^ ^^H 


^k' '^ c^ a # P ( ' IIHI^li^H^ ^^H 


T^. "-'^'vj* '^ '^^'i^'^ll^B^^^llH 


pS* ■, 'i^t** .^. ; .riliPiP^IB 


^'^j *, '^^sM^S^^^ ImB 



. .■ .IfiKS 

^■*'.''';JME*- '' 

i 1 

I^CiIBBI' ft 

W>^.J!'-' '<$ 

I "^^sK^ 



'j^'. jSWB-j' 


r' % 


^ty0K' ' .■; 


I^JTy: w||CTraW^B!j»^ttyW^-^^J|^MM^^H 




wjr J' .al^^QgRV^^^BSHRwl^iFa^flBar^'T9SS99^09^^H 

KjMW^^jf^aaMBBBB'Bw. i^^ ^Jv" hHSUb^SsBkB^ 



culturists have dotermiiKMl ])v exporieiice to l)e most favorable for the set of spat, 
usually almut the niichlle of April. Their arrangcMuent and mcMle of replacement 
Hr<* as follows: Shibi are brought to the oyster-grouuds in skiffs and deposited on the 
flats as they beeomf^expose<l by the receding tide. The eulturist will have had the 
boundaries of his concession staked out, and he has but to construct his fen(*es of 
shibi as 4ui<»kly as possible to take advantag«* of the hours of low tide. As a time- 
wiving device, h<» has already ha<i the ends of tin* shibi shar|)ene<l so that they can 
be thrust deeply (on<*-quarter or one-fifth tlu*ir entire length) into the soft Ixtttoni. 
Should thelM»ttoin prove hard, howov^M*, holes an* fii-st mad(* for the shibi by nutans 

•^ ^fc^^"" 

^^ ^B ^^ m 



m ^m ^^ 


■ """~ 

^ " 

■•^ ■ ' 


»^^m ^mm 





• • 

• • 



• • 

• 1 



• * 


• • 


:r ^ 


■ • 



• • 






• • 

■ I 




• fl 

• • 


• 1 

• 1 

• 1 







t ' " 



• 1 












t • 


• * 


• • 

■ • 

• ■ 


• a 







• 1 

• 1 



• I 





• I 





• 1 

• 1 























•1 ■ 





• 1 








•c ■ 


t « 


• • 



• 1 









• 1 


• • 


• 1 


• 1 







• I 


• 1 







h ■ 



• 1 




• • 








• • 

• 1 


• • 

• • 

• < 


• t 




1 1 ■ 








■ 1 



• 1 






• 1 

• 1 





• 1 




\\ i 







■ • 


■ 1 

• 1 

• 1 





• 1 


• • 



















• • 

• 1 
t t 













t i 













• I 
















• I 




• 1 

• I 

• I 


• • 




• • 






Si ■ 


• • 
















• I 

• I 










a • 













1 1 








• • 

• • 











• 1 












• 1 

• I 



S 1 


• I 


■ • 

• • 









f • 













• a 



• • 

• « 

• • 







t a 




Il ■ 



• • 


■ • 

• I 


• t 



1 1 






• I 



• • 



• • 








• • 


• 1 

• • 

• i 




• 1 




• I 

• 1 

• t 

• • 






t • 



■ I 


^^B^HMBB m 




.^ — 



..«— . 

.— il— - 



■^«i^«M m 





^i» ^^»^^ 







^^ ^^^ ^^ ■ 




\^m ^MMkSV 


^W ^» ^B ..• . 

■k ^m^gm ^^ ^B 




^^ ^m «^ « 





Fii; !•». Diairmin nf an 'lyst^T funn in whicli an* r<»mbine«l thi> l<>nL:itii<1inal ;ind tninsviT*^* ni«Mii*s of 

arrau^rintr th** )iuinl><M» colltM-tor-. Ku'^it^n. 

of an iron-shod pick, shown in tig. .*{. This thr workman sometimes pre.s.^<'S <lown 
with his foot, honcr the lateral support near the head of the implement. 

Aeeording to the usual type of oy>ter farms in Kaida, th(* nuiin boundaries are 
planted nearly parallel to the tide marks. Similar rows of shibi are next thrown 
out in tlKMlirection of the middle lin<'ot' the park. ( Figs, s, '.•.) They do not. how<'ver, 
meet, but leave* an oi)en median space passing through lik(» an aisl<». Thus on eith<»r 
side of the main aisle of the ovstei' ]>aik there are rows of transverse* allevwavs, (»ach 
alK)ut ♦> fe(*t wid(*, which te^rminateoften blindly at th<* main fenceof the j>ark. Details 
in arrangement are given in figs. 4, '>, <». and 7, fig. <*• showing a horizontal projectiem 
of the shibi of fig. 5. I'he fences and partitions of shibi stan<l about waist high. 




Fio. 11.— Diagram of a small oyster farm of labyrinthine 

pattern. Ondo. 

Such an inclosiire in Kaida Bay is practically the whole stock in trade of the 
culturist, for in it spat is taken and the oysters ^ow attached to the bamboo until 
the time when they are sent to market. From the permanent character of the park, 
therefore, the culturist has been led to ingenious arrangements by which the shape 
of the park can he retained and the oysters grown, j^oung and old, side by side. To 
accomplish this each fence is usually formed of a double I'ow of shibi, as indicate<l in 

figs. 8, 9, of which one row is of the second 
year and the other a new one. The latter 
is often added as the older row is removed. 
The fences in this locality are made of 
the more delicate species of Immboo, 
**hachiku." The stalks are trimrae<l in 
lengths of about 5 feet and are inserted a 
foot deep in the bottom. The distance 
between the double rows of shibi is inten- 
tionally narrow, so that their opposing 
l>ranches can interlace. (Fig. 7.) 

The foregoing description refers to 
the commoner type of park in Kaida. 
Occasionally some are seen of a more 
complicated pattern, as in fig. 9, and, 
rarely, some in which the double rows of shibi are placed parallel instead of trans- 
verse to the main axis of the park, as in fig. 10. An excellent example of this type 
is pictured in plate 3, from a photograph which was taken at the end of a season, the 
shibi showing well-grown oysters (from park at Tanna). Rarer still is a form in 
which a labyrinth-like arraugement of the hedges of shibi prevails (fig. 11), or even 
a concentric pattern (fig. 12). In all these forms, however, the arrangement is such 
that many eddies will be formed about and within the rows of the shibi, since these 
eddies have been found con- 
ducive to the attachment of 
the young oyster. 

In the foregoing types of 
parks the visitor notes that 
the height and strength of 
the shibi, their simple or 
branching character, to- 
gether with the closeness in 
their arrangement, vary 
somewhat widely in different 
localities. Occasionally an 
arrangement which alternates old and new shibi in the same row is foun<l U) he 
adopte<l mlvantageously. 

At the close of the cultural process, that is, at. the end of the second year, it 
remains only to remove the marketable oysters from the shibi. This is done during 
the favorable tides, the culturist using a pick-like instrument (fig. 13) with one hand 
and seizing the shibi with the other. For the protection of this hand a curious 
heavy but open mitten is used, figured in fig. 17. After the oystera are detached 

Fiu. 12.- Diagrram of an oyster farm in which the t^lle(*t<>rs are arraniced 
in c'irt^ular and concentric order. Middle clear Hjiat't^ umhI aM a living 
ground. Itxukushima. 


they are tAken from the ground by means of the rake shown in fig. 15, placed in 
baakftw, fig. 18, and carried thence in the usual oyster boat often to the mouth of 
Bome adjacent river, where the^vare thrown outand rakedover. Bythe latter pro- 
cess, "drinking" the oj-sters in fresher water, they increase in size and become 
cleaner, a process, by the way, quite similar to thBtempl«ye<lin France, in England, 
and often in America. 

ciBine wi'll-gTi>wn oyBli-rs rmm tl 

At Kanawa, an important cultural ground, a similar method to that of Kaida 
Bay is employed. The cultural area is not large but it is ver^- productive, and here 
they have found it profitable to plant shibi in close rows at right angles to the coast 
line, as shown in fig. I'.l. 

The oyster-grounds of Kusat^u are the best of those situated west of Hiroshima, 
but all of these, and of this entire region, are essentially the same, as far as cultural 

H. -Oyster nke. 

(B/u^-jmim, oiwd 


«,A.m-i«m/. aaed 

oysters, I. e.. Miirinit tbem si 
k3t» Hhell nurgln 

Pin. m.-Oyatcr rake, n-iltua-i, D»pd t-i rullw-t niArketilile oyBleni frum the liring (p-cinnrl, 

methodsareconcerned. They extend along the western coast 7 or more miles from 
Hiroshima, at points indicated on the map (fig. 1). In these localities oysters are 
cultivated at greater depths than in other waters of Aki, for it has here been found 
that under the deeper conditions the shellfish continue to increase in size after the 


seeimcj year, unlike those at Knida. Thcivforc, at Kusiitsii, to the en<l i>f growing 
a lanier oyster, tho cniltiirist divides his ]>8rkH iiit.o three t-lasst's: thofw.' of rthiiUow 
wat^T (largely for spat <;ollci'ting), early rearing, and deep wat«'r (for lale rearing). 
The cult mill <'Oiic*es.sioiisa<'<'onlingly have enme 
to Ik' arranged with their long axes a1 right 
angles to the shore, thus pi'oviiling a range of 
wat<'r iMissing from shallow water to deep, each 


winter l.Oiid. The density of the wat^r risen 
gradnallyaiKl attains alxtnt ]M-25 in thedee[>etil 
zone. Aoeoi-dingly, the shallowest region in 
eaeh park usnally Iweonies laid ont in a zone of 
eollectoi-s, or shil>i-))a, and i-esemhlesHoinewliat 
Kaida Hay. In the next ami deeiH-r ground 
are the rearninged andoyst^-r-ltearingshilii (of 
the shallower zinie), loya-lta. and in lhedee|K'st 
water are the typical oyster In-ds, or iiiiire-ltu. 
Of <-oni-se sueh an arrangement is sonielinies 
motlilied, sinee priwtiee demonstrate^ that the 
liM-al eonditions of an oyster jiark, i. </., water 
<-urn'nls, ai-e apt to warrant widely difTerent 
('() In the shallow wme an arrangement of shibi in lines imrallel to the shore 

is theeommon one. Uelweeu the rows are intervah* of about 4 feel, the park in 

UiiHeaseivs.Mnblingone of I he eonnnon tyi«- of Kusal.-fu (lig. 8). UfU'n, however, 

the shibi are shorn of their branches and 

plaulod like eanes (;i U> r, feet in length) 

in elose-sel rows. Sueh, for exatiil»le, are 

the shibi jihotographi'd in plat<' 'i, .1, /.', 

and O. The first (if these, .1, has had the 

oysters attached to it for alwut a month; 

the se«t»nd, fi, for about ti months, anil the 

thinl, (■', for al>ou1< If* months. It is at 

the last iieriod that the masses of oyslei-s 

etime to sepaitvte somewhat tvadily from 

theltivnilKK). In these parks the arrange- 
ment and treatment of the shibi varies 

greatly in aeeonlanee with loeal eondi- 
tions. In rapid enrrents, whi<-hdis1in^nish 

the njgion of Kusatsn, short ami branehle 

of the strong and short-jointe<ls|)eeieM of baniluxi, "madaki 

is f ro<iuontly in clumps, or t-<iya, iis imltcated in ligs. '20 to 

4 shihi 

I, Ihese. loo, miwin 
riieir arrangement 
( )f these the 1«j-a 

11. U.S. F.C. 1901. ITgfK.pig. 


of figs. iiOftiiil n^l an- si>e<-iHlize<l Tor shallow water, tiiid tliiit of li}r. 2:! to h mpid 
curn-in. Ill Kiicli cliiiiips bmnehiiig shilii fn-tiut-iitly (H-ciir, nii<l in tfaiH event the 
tips of till' Ktalks are more apt to diverge Ibaii in Kaida fcontnist fifis- --t and ^i). 

Flu. JU. Iii:i;:mm uCoysiur tnzm in wbK'h slubiari.' jiUnicil iu ruws ixkniUi'l uud ji riglii ttiwlw !•• Kbure lin.'. 

Hiiotlii-i- ailaptation li> more mpid cnn-eiil. In (jreiifral tlie lutses of tlie coinpuncnl 
tttiibi aif iiii[ilaiiti-d alxml a fnol. Tliiis iiiade llie 1<>ya nMiiaiii in position for about 
two years. 

((') In the ue.xt de<.'|M'r zone, tyjiieal liiya-lia. tlie jrniniiiiifr of shilii ln'c<inieR 
more nia.ssire, and when at low iid<' this ivyiim is o.\posed (lit In l-"i feel is ilie tide 
tall) one sees tlieiii in lon;i; rows wliii-h sii}r;rest diiiiiiiiitive liaycoi-ks. An e.\(i'lleiit 
idea of sueli toya-iw is had in [ilaie 5. C'loiier inspeettun shows thai oaeli toya isi 


in nearly every e«8e a complex of shibi of many ages. Transplanted shibi, with 
oysteiTS of one or two years' growth, usually form the nucleus of the cluster, and 
around them ai*e planted concentric rows, one, two, or thi'ee, of branched shibi, old 
and new, to the end that all ultimately bind or mat themselves into a living, springy, 
1 cone-shaped mass, well su ited to resist currents or storms. ( Fig. 25 and PI. 5. ) Such 


• • •% •• •• •• 

••• ••• ••• ••• ••• 

• • •• •• •• •• 

• • •• •• •• •• •• 

••• ••f ••• ••• ••• ••• 

#• •• •• •• •• •• 

• • •• •• •• *.. • 

— ♦ ••• ••• ••• ••• ••• 

• • •• •• •• •• 

«• •• •• •• •• •• 

••• ••• ••• ••• ••• ••• 

•• •• •• •• •• •• 

Fio. 21, — Arrangoment similar to flf<. 811, bat in oltliiiue ordf^r. adapt^Ml to a Homowhat iiioro rapid <*urr«*nt. 

toya are matle, or remade, lowaiil the end of each spawning season, /. «'., during 
the end of August or early part of September. Tliey are then pulled to pieces (in 
this work the rake shown in fig. 14 is used), and from each bamboo there are shaken 
and broken off the oj^sters which are least securely attached. During this process 
the shibi found to be still useful are put aside to form the nucleus of new toya; 

• • •• •» •• •• •• 

••• ••• ••• ••• ••• ••% 

• • •• •• •• •• •• 

> • •• •• •• •• 

• • ••• ••• ••• ••• • • 

»• •• •• •• 

•• •• •• •• •• •• 

• •• ••• ••• ••• ••• ••• 

• • •• •• •• •• •• 

Fin. 22.— Arrangement similar to flg. 20, but adapted to rapid current, ruwH of coUoctom 8i*i)aratod 

by intervals of K or 10 feet. 

the gleanings, twigs, detached oysters, and all, are now raked up and carried to a 
third locality, the living grounds (ike-ba). Here in swifter current the debris washes 
away and the oysters remain, the shells usually becoming clean in the meanwhile. 

(c) The living ground, or ike-ba, of which a good example is pictured in plate 4, 
is generally located in a zone of deeper water. It has a clean, gravelly bottom, 



which ill motif vases Iteeoines exposed only nt tlie lowest tides. Here the oysters nre 
broiinht together which liiivo Iiecome (let»ehe<l from shibi in various parts of the 
farm, ami thus they are relmiied in elassified l»eils imtil they have attained their 
second year's growtli. During tliis time the eulturist 1ih« l)ut to keep them well 
spread out and to see tliat the beds ai-e Itept clean. Always at lowest tides, and 
sometimes as oft«n as fortnightly, the laborers give the oysters a vigorous raking 
(using for this purpose the short- toothed rake shown in fig. 14), which scatterH them 
about the bed, removes foreign bodies, and, l>est of all, gives the shells a better 
sliaiM- and a firmer rim, for in this treatment the delicalo, cnticular outgrowths of 
the shell are removed and a more symmetrical growth result*. Of especial impor- 
tance is the pi-ocess of raking in cases where the oysters are sent directly from the 
Uviug ground to market; for it has been found (here as elsewhere) that those oystera 

tatnlimi rollertora mmntft-'t » 

fter thf fwbkiD ™imiion in K 

[wtau. Tb..»hiT> 

fwt nliuT^ tlw buttiim and th 

oir tLp- ciivew; the <,liuni« »'.t.*tB 

whose shell rims are strong and accurately fitttHl together fai'e better during trans- 
portation and in the market. Those with delicate aud brittle margins soon suffer 
injury and lose their fluid through leakage and evaporation. 

In some farms, on the other hand, it is maint^iined that oystei-s of different 
sizes should lie mixed on the same living ground. l*"or it is claiinetl that the 
young oysters grow better side by side with the older ones, and even that if the 
more iwrfectly grown oysters of different grades can lie mixed together during the 
process of raking, the lietter will be the general output. 

(if) The final stage in the culture of Kusat-^u takes plat^'c in maturing grounds, 
or miire-ba. llei-e the larger oysters of the second year's gi-owth an- laid down, 
transplanted from the one or more ike-ba of each establishment. I'sually tliev are 
in the deepest water cultivated, *'. e., in water a few feet in depth at Uiwest tides 
up to water of 3 or 4 fathoms. 1 was tuld that in one luiire-ba oysters were vulti- 


vaU-it siKxt-ssl'iilly ill wiitit'i- of 5U fi-fi- Ejm-Ii Canii in KiiMitKU has itw M'liarnto 
miiiv-bu, and tlnwc" iKljoiu mn- huuIIut. toniiiiig a t^tontiiiuous ziniw in (ieepoi- waU-r, 
ea<rh iiiiirt--l>a ilesigtmtHi by a iiiiinlHT. In the rthallowoHt ]>oi'tiiins lli» oyHtei-N »■'» 
iiHiially pniUHit^-tl H;;;iiiiist, iliH]>lticeiueut and invasion uf muil li.v ntt-aiisor low fi-iicoM 
Hrraii|i;t^<l with \vin{;;-likv <-xiMinsioiiH, an 8liown in };r<iun<I plan in I'm. l'i>, and in <lt>tail 
iufig. -7, This devitw lias Iwt'iidcvclopofllarjwly in vii>w of thu storms (ifthiityplumii 
ueason. The uffout of tlic- uiaturiu;; gi-omid i.-* to givo thi.' oysters greaUrr n'v/Ai aad 

weight, and to givi.' tlie iiioi*t a whili^r color. The tinislicd pnxlnci I'oscinhifs i-loscly 
a woll-gi-owni oyster nf Long Ishind Sound. Markrling Ijikcs iilacc nftiTlin- oyslci-s 
have reinaiiKHl fiMiii six mouths Hi a year in thu deeper gi-oiind, making ilieir lotjil 
age ahoiit thwv years. 

Each oyst«r farm lias iu separate tioiises, situnt<-d usually on llin luljaeont 
siiore, and tlie details ia handling and packing iln. oysU-rs appear to In- closely 
similar lo those of Oontinenhil Kiinipe — baskets. Mocks, rakes, arrangement, and 



storage. I may note in passing the curious **knife,"a combination of kuife, mallet, 
and lever, with which oystei's are opened with surprising rapidity. The modus 
operandi is shown in fig. 2, (mge :J1. 


In this locality (cf. fig. 1), finally, oyster-culture has Ix^en developed on more 
special lines than any whei*e else in the East. As at Kusatsu, the industr>' embraces 
three distinct branches — (a) spat collecting; (b) rearing the young, and (c) maturing 
and fattening theoystei-s for market. But, unlike at Kusatsu, these three branches 
of the industry are carrieil on, not on the same shoi-e i-eaeh, but at points widely 

Fh; i'».— Gnuind plan «if h uiound toyn of r«>lle«-tors. Shibi with wt»n-jfrowu oyHtcrs are 
indicaU^l by the bUu'k ^i^tts within the two <>in'k>b of bruuchin^jr Hhibi. 

separate. In othei- words certain sfH^cial tracts are taken advantage of in collecting 
spat; others are s|M»cially arrangi»<l for early rearing; others, in turn, for maturing. 
In these regards oyst4»r-cuUure at Nihojima resembles closely that of the coast 
of Brittany or of IIollan<l. The details of the management of the farms are in 
essentials, however, like thosi* previously described. 

(a) Sptd coUvrt'nuj, — The sp*it is collecte<l in very shallow water, less than a 
fathom deep at the usual tides, temiierod considerably by incoming streams. The 
specific gravity is sai<l to nirely exce<Ml 1.017. In such a i-egion shibi are put in 
position, usually in very clos<» order, at the l>**ginning of each spawning season, say 
from the middle of April to the middle of May. After a period of about three 
months the entire mass of shi1)i will be uprooted and transplanted, sometimes a 
mile or more, to a hx'ality Ix'tter fitted for rearing llie young oystere. Tliis trans- 
portation, 1 was told, is the most diflicult part of the work of the culturist of 
Nihojima, for the minute oystei's are, as everywhere, peculiarly liable to injury; 

F. C. B. 19W-« 



carele88 handling will destroy great numbers of the delicat>e shells; hasty {mcking of 
the shibi on the scows, whereby the branches are allowed to rub together, is another 
palpable source of loss; drying, direct sunlight, changes of teniiHirature, are all 
deadly, and, above all, severe thunderstorms — the latter, according to my informant, 
causing death by /ear. I suspect, however, in the last regard, ihat a fresli, cold 
shower bath is more apt to be a moving cause, although I was assured that, the scows 
are covered with the ever-present Japanese oiled pai>er to guard against such a 

{b) Rearing the youmj oysters, — The shibi, covered w^ith young spat, ai-e now 
arranged in toya-ba, like those of Kusatsu, but closer in arrangement, and usually 
of many varieties. Here they remain from one to two and a half years. In the 

Fia. 26.— Diagram of livinir ffnmnd of two to three yoar old oyiitcrH. Tlit^ HiiotH rt*iirom>nt oynt4»rH and the dark 
lineH art^ rowH of shibi plaetMl ho as to provide shelt^srs against currentH or Htorins. The arrow denotes direc- 
tion of strongest current. 

case of the older and n^arranged toya a long, mound-like tyiM* is <*<)mmoner t.han 
the circular ones descril>ed above. As a rule the toya are covered with water saive 
at lowest tides. 

(c) TxUer rearing; Maturing, — The living inclosures in dei^per water, ike-ba, 
correspond to those of Kusatsu. They <M)ntain the oystei's whieli have lK*en sejwir- 
ated or are readily separated from the brush of the toya. A similar pi-ocess prevails 
of raking the oystei*s I'oughly, and I wiis shown some sh<*lls of the ohler oysters 
from this region which were of very i*egular shajK'. At favorable tides, further- 
more, the oyster beds are cleaned and the boring whelks — Purpura rlarigera Kaster 


ami Jiapaufi iM'Z'xir L. — are reiiKtved. StiirlUli are not tfoublesoine. Further traiiB- 
plaiiliii;^ takeH plm-e, UHunlly at the end of llie iseeoud seiMon, and in kUU deeper 
water the «yst«;r attaitis finally its marketable size. 


All eiiltivalile grounds, wlietlier fur oystei"s, other i^heUfish, or 8eawee<l, are — in 
Aki, at least — the proi»erty of the iirefceliiro and can be neither sold nor sub- 
rented. All cultivjible tracts are surveyed, the hits tending to decrease in size as 
the estimated value of llie property inereases. Tlie mode of lajing out concessioDB 
can perhaps be best nudersUMxl l>y i-efei-ence to plate 7, copied from a Government 

chart. The farms ari- rented auction fashion to the hii^hest bidder, and the Ijeiianis 
have the privilege of n^newing llieir lea-^es indefinitely at the original reatalu, a 
privilegi', however, which can not l>e used speenlatively. When at the termination 
of a least! the proiMily jmsses again into the hands of the prefecture it is at once 
advertised and rerented. In addition Ut the yearly rental the proi>erly is subject 
to a small local ta.'i upon the total area of cacli farm, and to a chai^ of 1 per cent 
of the n-ntal to cover the e.\penses of administering the oyster-cultural bureau of 
the prefecture. Tliis series of taxes impresses the stranger as formidable, until he 
learns that it does not repivsent an ac<-umulattou of taxes, but rather an itemized 
statement as to the apportionment of public funds made thus in accordance with 
local euslom. Henlal of concessions from the prefecture, it should further be said, 
brings with it certain privileges in marketing the oysters in Osaka. 




To what degree the Japanese oyster would flourish on our coast can be deter- 
mined, obviously, only by experiment; and if experiments are to Ik? made, the 
following suggestions seem to me of some practical value : 

On the Pacific coast, on grounds which have been found especially favorable for 
the reproduction and growth of our Western species, O. califoniira, consignments 
of Japanese oysters may be planted — in the north with oysters from the Hokkaido 
beds, preferably from the region of Mororan or from Akkeshi Hay, and in the south 
from the region of Hiroshima. In this way similar conditions of temperature will 
be obtained. To fulfill a second favorable condition an effort should ha made to 
secure oysters from water of nearly the same specific gravity as in the chosen 
American localities. There would be, I fancy, little difficulty in the matter either 
of securing oysters from equivalent localities or of transporting them. Through the 
Imperial Bureau of Fisheries, under the able headship of Dr. K. Kishinouye, one 
could promise, a priori, prompt and efficient aid in getting in touch with the Japanese 
oyster-culturists whose establishments are known to be favorably located. And 
with fast freight service from Yokohama the oysters could be transport-ed with a 
minimum of loss, as similar exportation {e. j/., of Ameri<*an oystera to England) 
demonstrates. In a case of this kind, however, extra precautions would not be out 
of place. Large oj'^sters should first be selected, and, preferably, treated with the 
raking process on ike-ba for a few weeks. J5y this means the shell margins will l>e 
thickened, and thus the oysters will lose a minimum amount of fluid during 
transportation. Care should also be taken to pack the oysters each with the more 
concAve valve downward and to mark the cases so that during shipment they shall 
be kept in the right position. Other details — if not indeed the above — can safely 
be left to the skilful Japanese work people. 

A further suggestion is that the shipments be made during the months of Feb- 
ruary and March, to the end that the coldest season on our coast would be avoided, 
and thus the oysters would have the chance of becoming in good condition and 
somewhat acclimated by the following winter. Moreover, at this season of the year 
it has been found that the oysters have laid up the maximum supply of nutriment 
against the breeding time, a supply which could be used as reserve nutriment dur- 
ing transportation. 

No experiment, however, could be regarded as a fair one, I believe, which did 
not deal with an adequate number of the imported oysters. If but a few oysters — a 
score or two — are laid down in each experimental locality they may be lost through 
accidents which would not befall a larger number. For in oyster banks there is 
certainly strength in numbers, an cecological feature in oyster-culture which 
governmental regulations in France, Holland, and Germany keenly recognize. By 
this numerical factor it appears a true paradox that one thousand planted oystei*s 
have more than ten times the chances of survival than a hundred. As oysters ai-e 
cheap, it seems to me better economy therefore to make the more convincing 

To what degi'ee is it practicable to introduce the methods of Japanese oyster- 
culture into the United States? This, I take it, is a (juestion which can be answei-ed 





This paper sets forth the experiences of nine seasons, beginning with 1894, 
during which I have had charge of the black-bass work of the Michigan Fish Com- 
mission. The work was begun at Cascade, Mich., and after four seasons was 
transferred to Mill Creek, where it is now carried on. The methods of pond culture 
finally adopted are based on a knowledge of the breeding habits of the fish under 
natural conditions. The account has reference to the small-mouthed bass, unless 
the large-mouthed is specified. 

In studying the habits of the bass, it is necessary" to distinguish the males from 
the females; ordinarily this is not possible except by dissection, but at spawning 
time the female is distinguishable, even at a distance of 10 or 20 feet, on account of 
her distension with eggs, and this makes it passible to determine the part taken by 
each sex in nest-building and the rearing of young. 

The nests of the black bass are built by the male fish working alone. The 
small-mouthed bass prefers a bottom of mixed sand and gravel, in which the stone 
ranges from about the size of a pea to that of one's fist. As the si>awning season 
approaches the males are seen moving about in water of 2 or 3 feet depth seeking a 
suitable resting-plai*e. Each male tests the bottom in several places by rooting into 
it with his snout and fanning away the overlying mud or sand with his tail. If he 
does not find gravel after going down 3 or -1 inches, he seeks another place. Having 
found a suitable place, he cleans the sand and mud from the gravel by sweeping it 
with his tail. He then turns over the stones with his snout and continues sweeping 
until the gravel over a circular spot some 2 feet in diameter is clean. The sand is 
swept toward the edge of the nest and there forms a few inches high, lea\ing the 
center of the nest concave like a saucer. The nest is usually located near a log or 
large rock so oh to be shielded from one side. If the bank is sheer and the water 
deep enough, the nest may be built directly against the bank. If possible, it is 
place<l so that the fish can reach deep water quickly at any time. 

During nest-building no females are in sight, but when the nest is done — and 
this takes from 4 to 4*'^ hours — the male goes out into deep water and soon returns 
with a female. Then for a time — it may be for several hours — the male exerts 
himself to get the female into the nest and to bring her into that state of excitement 
in which she will lay her eggs. If she lies quiet, he turns on his side and passes 
beneath her in such a way as to stroke her belly in piissing. If she delays too long, 
he urges her ahead by biting her on the head or near the vent. If she attempts to 
escape, he heads her off and turns her back toward the nest. If, after all, she will 
not stay in the nest, he drives her roughly away and brings another female. 

a Read at the thirty-first annual meeting of the American Fisheries Society. 



Some 15 to 30 minutes before the female is ready to enter the nest and spawn, 
her excitement is made evident by a change of color. Ordinarily she appears to be 
of a uniform dark olive or brown above, changing to a light green l)elow. The only 
markings readily seen are four stripes on each cheek; but in ideality the sides of 
the fish are mottled with still darker spots on the dark-olive background. The spots 
are arranged so as to form irregular, vertical bands like those on the percli, but 
these are not usually visible. Now, as the excitement of the female increases the 
background becomes paler and finally changes to a light-gi'een or yellowish hue, so 
that the spots and bands stand out in strong relief. The whole surface of the fish 
becomes thus strongly mottled. This is a visible sign that the female will soon 
spawn. The male undergoes a similar but less pronounce<l change of color. 

Soon after this the female enters the nest and the male continues to circle about 
her, glide beneath her, and to biti* her gently on the head and sides. At times he 
seizes her vent in his mouth and shakes it. When this has continued for a time 
spawning takes place. The two fish turn so as to lie partly on their sides with their 
vents together and undergo a convulsive fluttering movement lasting 3 to 5 seconds. 
During this time the eggs and milt are extruded. The circling movements are then 
resumed to be interrupted after a few seconds by spawning. This alternate circling 
and spawning continue for about 10 minutes. The male then drives the female 
away, biting her and showing great ferocity. She does not return. 

The male, and the male only, now continues to guard the nest, fanning sedi- 
ment from the eggs and repelling enemies. At GG° F. the eggs hatch in 5 days and 
the young fish swarm up from the bottom in 12 to 13 days from the time of hatching. 

Henshall in his "More About the Black Bass," published in 181)8, ciuotes, with 
approval, Arnold's observ^ations to the effect that the nests are built and then guarded 
by the female. The Manual of Fish Culture, published in 181)7 by the Unite<i 
States Fish Commission, speaks of the nests as being built by the mated fish, some- 
times working together, sometimes separately. These seem to be the latest published 
observations, and are not at all in accord with my observations in Michigan. 

Shortly after the young small-mouthed bass rise from the nest they scatter out 
over a space 4 or 5 rods across — not in a definite school with all the fisli moving 
together, but as a loose swarm, moving independently or in small groups. This 
makes it impossible to seine the young fry, as upon the approach of the seine, 
instead of keeping together, they at once scatter and escape the seine. The fry 
may be at the surface or on the bottom in weeds or clear water and are attended by 
the male until they are 1^ inches long. The swarm then gradually disperses and 
the young fry, which were previously black, take on the color of the old fish. 

The breeding habits of the large-mouthed black bass are similar to those of the 
small-mouthed, but differ in some respects, which are of importance in j)ond culture. 

1. The nests of the large-mouthed are not made on gravel, but by preferc»nce on 
the roots of water plants. These are cleaned of mud over a circular area, and (m 
them the eggs are laid. As the eggs of the large-mouthed bass are smaller and 
more adhesive than those of the small-mouthed they are apt, when laid on gravel, to 
become lodged l>etween the stones and to stick together in masses, and are then likely 
to be smothered. When laid on fibrous roots of water plants this does not occur. 

2. The young large-mouthed bass remain together in a compact school very 
much smaller than that of the small-mouthed, and the fry usually move all in the 
same direction. This makes it easy to seine the large-mouthed fry when wanted. 



Ponds antJ stockfish. — After some exjjerimenting all our ponds, lK)th for stock fish 
and f ly , are built on the mo<lel of a natural jwnd. There is a central deei>er portion 
or kettle about G feet deep, and around the shore a shallow area where the water is 
about 2 feet deep. The lK>ttoui is the natural sand, and water plants are allowed to 
grow up in the ponds. All iK)nds are supplied with br(X)k water, and silt from this' 
furnishes a rich soil for the aquatic plants. The water of these i)onds contains 
Dciphnia, Bosmimi, CoriXy and other small aciuatic forms in great numbers. These 
furnish fiKxl for the bass fry. The ponds run in size from 120 feet by 190 feet to 
100fcM.-t by l(K)feet. 

At fii*st w(» were unable to fwMl the stock fish on liver, but after a time we found 
that by cutting the liver into strips al)out the size and shai)e of a large angleworm 
and by throwing the strips into the water with the motion that one uses in skipping 
stones they wriggle like a worm in sinking and are then readily taken, llie liver 
must l>e fresh. If bass are ftnl on liver alone they do not come out of winter 
quarters in gcKnl condition. Of eleven nests made by bass thus fe<l only three pro- 
duced fry. Although eggs were laid in all, they seemed to lack vitality, owing to 
the poor condition of the parent fish, and in eight of the nests the eggs died. 

In onler to bring the fish through the winter in good condition it is necessarj' to 
begin fet^ding minnows in St*ptemlK*r and to continue this until the fish go into 
winter quarters. The Imss eat minnows until they go into winter quarters, after 
which tliey take no foo<l until spring. The minnows are left in the ix)nds over 
winter, so that the bass, when they come out of winter quarters, find a plentiful 
supply, which lasts them until the spawning season. At this time the minnows are 
seine<l Innn the pond, as their pivsence interferes with the spawning. Before this, 
however, some of the minnows have spawne<l, and their fry later serve the young 
bass as Uhh\. Bass fed in this way come out of winter quartei-s in fine ccmdition 
and their eggs are found to l)e hardy. 

Ariific'ud ftrtilization. — During the tiret two or threi» seasons of our work 
numerous attempts wei-e made at artificial fertilization, but only twice with success. 
On one occasion the female was seine<l fmm the nest after she had lx»gun to s^mwu. 
She could then l>e readily stripjxKl. Tlie male was cut o\ye\\ and the eggs were 
fertilized with the crusluMl testes. About 75 i)er cent of the eggs hatched on a wire 
tray in running water, the eggs l>eing fann(»d clean (»very day with a feather. 

In tin* second <*ase the fish wei^e S4»ined while spawning, and it was found that 
in the case (»f one female pi-essure on the alnlomen (*ause<l a i'c*<ldish papilla to 
protrude from the vent. This had the api)eamnce of a membrane closing the vent. 
It was pinche<l off, and the female was then stripped readily and the eggs fertilize<l 
and hatched. 

Ptfud ridfure. — Having abandoninl artificial fertilization, our attention was 
turned to pond cultur«\ and this we have carried on for about six years. Our earlier 
ponds not furnishing natural spawning-grounds, we construct4»d alongside each of 
the large ])on<ls six smaller iH)nds for use as spawning-ponds, each about 10 by 24 
feet, 10 inches dt^'p, with gravel iMittom, and connected to the central pond by a 
4~foot channel. 

The fish entered these and spawne<l. In one case we had eight nests in a single 
pond of this sort. Where so many nests were made, usually but one or two of them 


came to any good, the others being destroyed by the fighting of the males. Ordi- 
narily but one or two nests were built in each spawning pond. The male first to 
enter and begin the construction of a nest generally reganled the whole pond as 
his property and held it against those tliat tried to enter after him. On one occasion 
the male thus holding the pond was attacked by 10 or 20 other males at one time 
and after a long struggle was killed and his nest destixjyed. 

The attempt to use small spawning-ponds was then abandoned and all the ponds 
were made of good size and with a central kettle and shallow shore area, as alremly 
described. The problem now was to prevent the fighting of the male fish and the 
consequent destruction of nest and eggs. I finally hit upon remedies for what 
seemed to be the two chief causes of this fighting. I had notice<l that in the natural 
wat^r the nests of the small-mouthed bass were frequently' built against a stone or 
log, so as to l)e shielded on one side. When so built the nests might be quit4^ close 
together, as near as 4 feet, and the fish did not fight, because they did not see one 
another when on the nest. On the other hand, if a bass nest wjis built wheix* it was 
not shielded the bass on that nest would prevent any otlier bass from building 
within 25 or 30 feet of him. It occurred to me to try to construct artificial nest-s and 
shield them so that the fish on the nests could not s(h> one another, phicing the nests 
so near toget.her as to fully utilize the pond area. 

In the spring, bi^fore the spawning setvson oj)ened, tlie ponds were drawn down 
so as to expose tlui shallow terrace along the shore. The terrace was tlien cleaned to 
a depth of about 2 inches of sediment and vegetation which luid accumulated since 
the previous summer. Rectangular nest frames of inch boanl wei*e made 2 feet 
square and without bottoms. On two adjacent sides these frames were 4 inches 
higli, while on the other sides tlu^y were 10 inches high. The}' were set where there 
would be al)out 2 fe(»t of watcM* when the pond was filled, and so placed that the 
corner formed by the junction of the two lower sides j)ointed to the center of the 
pond, while the opi)()site corner, formed by the higher sides, pointed toward shore. 
The frames w(»re set directly on the bottom, not in excavations, and each was filled 
with gravel containing sand suitable for nest-building. A board was laid diagonallj' 
across the two higher sides and a heavy stone laid on this to keep the frame in place. 
The two higher sides form a shield on two sides of the nest, while the board across 
the top affonls shade. The frames were set in two rows al>out the iK)nd, parallel 
to the shore line. 

The rows wer<^ about feet apart and the nests in each row alK)ut25 feet apart, 
alternating with those* in the other row. There was thus al)out one nest to each 100 
square feet of suitable bottom, or in each area of 10 b}' 10 f(H>t. Wh(Mi the bass 
were on the nests no one was able to see any other and the fighting from this cause 
was practically (eliminated. The number of rows of nests may Ix* inci'eased to three* 
or four, or more where th<» area of shallow water is wide enough. 

Tlie bass selected these nests in preference t^) any otlier spawning-ground 
They (»leaned up the gravel and behaved in the nests in every particular as the\' 
would on natural spawning-grounds. The first time we tried th(*se shielded nests 
not a single bass made a nest outside of them, though there was plenty of good 
gravel bottom available for this purpose. 

As to the second cause of fighting: In 19(J(), when these nests were first tried, 
from 475 stock fish we obtained 315,000 fry aiid 750 fingerlings. In the season of 
1891 the output was very much less and there was considerable fighting among the 


fish. This remained unexplained till the ponds were drawn down after the spawning 

season, when it appeared that although the fish had been sorted, the number of 

male fish was c*onsiderablv in excess of the number of females, and these excess 

males, banding together, went about breaking up the nests of their more fortunate 

brothers. It is now the practice when setting the nests to seine out the stock fish and 

sort them, putting about 40 males to 60 females, thus removing the second source 

of fighting. 

During the present season from 493 adult fish we had produced 430,000 fry up 

to May 2G, and we believe that we can do as well every year. 

Up to the present year there have been two sources of loss incident to the wat^r 

suppl3\ The supply is a spring-fed brook, which runs over an open country befon» 

it reaches us. The water in this brook becom(»s ([uite warm on a hot, sunny day and 

cools off at night. The temperature thus falls at night sometimes Jis much as 13° F. 

and becomes as low as 40° F. This is disastrous, since when the temperature gets 

below 5U° F. th«' adult fish desert the nests and the eggs or young fry ai*e killed by 

the sediment. By watching the temperatui'e of the water and, when it approaches 

50° F., shutting off the supply until the water warms up, this difficulty is obviated. 

Since the ponds are well stocked with water plants the fish do not suffer from lack 

of oxygen when the wat^^r is shut off. Indeed, if the water did not leak out of the 

ponds, I doubt if it would be necessaiy to introduce any running water into them 

during the breeding season. 

The second difficulty with the wat^r supply is from sediment brought down by 
the brook after heavy rains. This sometimes accumulates over the nests so thick 
as to smother the eggs and drive away the parent fish. B^*^ shutting off the water 
supply whenever th(* water is much roiled this trouble is avoided. 

The water supply, however, must be kept fairly constant. If the level lowers 
more than about inches the fish leave their nests and the eggs die. For the purpose 
of maintaining a constant water level it would probi^bly l>e tost to have the ix)nds 
made with clay l)ottoms. The difficulties arising from roily water of variable tem- 
l)erature are, however, local, and would probably not be usually encountered. 

Handling the fry afier ih( ij rise from fhe nesf. — The small-mouthed fry have 
the habit of scattering into a large swarm when they leave the nest and it is con- 
sequently difficult to seine them when wanted. It is therefore desirable, just l)efore 
the fry rise from the l)ottom, to set over each nest a cylindrical screen of cheese- 
cloth supiH)rted on a frame of band iron, fi i-st i-emoving th<» wooden nest frame. 
The screen keeps the fry together. They thrive and grow within it and may be left 
there until one desires to ship them. The old fish stays outside and watches the 
screen. When this supply is gone other Crustacea nmy be taken from the pond 
with a tow net and placed insille the screen. The fry are removed, from these 
screens directly to the shipping cans, as wanted. 

Raising the fingerlings. — The water in one of the ponds is lowered, the old fish 
seined out of the kettle and transferred to another pond; the pond is then refilled, 
and the fry, now about one-half to three- fourths of an inch long, are put in. The 
water in the pond is thick with Daphnia and other Crustacea, and these do not get 
out when the water is drawn off. The fry feed on them and the supply is usually 
sufficient; but if it gives out, a fresh supply may be gathered from another pond and 
placed in the nursery pond. As the young bass grow they eat not only the Daphnia 
but young Corixay and doubtless other aquatic animals. 


In 1901, fry one-half to threo-fourths inch long were introduced into the nursery 
pond on July 12; on August 5 they were seined out and shipped, and were then 2 or 
3 inches long. They had had none but the natural food. In three mouths these 
fish, under the same conditions, are 4 to 6 inches long. 

I have spoken so far of the small-mouthed bass, and it remains to say something 
of the large-mouthed, with which my experience is more limited. It is less necessary 
to resort to pond culture with them since, owing to the habit of the fry of keeping in 
a close swarm, they may be readily seined from their natural waters shortly after 
they have left the nests. In cultivating them in ponds I use the shielded nests 
already described, but make the bottom of some fil)er, preferably Spanish moss 
bedded in cement, as has been suggested by Mr. Stranahan. This imitates the 
natural nest bottom and gives better results in our locality than the gravel nest. I 
do not place screens alK)ut the nests, since the young fry are so small that it is 
difficult to hold them with a screen, and since they may be readily taken with a 
seine when wanted. I allow the large-mouthed fry to leave the nestfl with the parent 
fish and seine them when wanted. 

Finally, I will sum up what seem to me to be important points in pond culture 
of small-mouthed black bass, the ponds being constructed, as is usual, on the model 
of a natural pond with a central kettle and shallow sliore region, well grown up 
with water plants, and supplied with lake or brook water: 

1. Fish should be so fed (with minnows) as to be in good condition in the spring. 

2. They should l>e sorted into the ponds in the spring in about the proportion of 4 males to 6 

8. Shielded nests should he used, arranged as already described — about 1 to eac*h 100 square feet of 
shallow water. 

4. The gravel in the nests should he carefully selected; it should contain sand and plenty of small 


5. Water on the nesting-grounds should be kept constantly at a level between 18 inch^ and 2 feet. 

6. The water temperature should be kept constantly l)etween 66^ and 75' F. (in our locality). 

7. Roily water should be, as far as possible, kept out of the ponds during the spawning season. 

8. Fish should not be disturl^ed until the eggs are hatched. 

9. The nests of the small-mouthed bass should be screened just before the fry rise from the bottom. 
10. The water should contain an abundance of natural food for the fry. 

The processes described are perhaps susceptible of improvements, viz: 

1. Special nursery ponds might be provided for rearing fingerlings. 

2. It is perhaps desirable to have the nest frames shielded on three sides instead of two sides, and 

made with a bottom; then when the fry rise from the nest, close the fourth side of the 
nest frame by sliding a screen into it. In this way it would not be necessary to remove 
the nest frame and put a screen over it, but the frame could be left in place and the open 
side closed with a screen. 

3. If the i)onds were made with clay bottoms, the water supply could be entirely shut off during 

the breeding season, if necessary. 

Contributions from the Biological Laboratory of the U. 8. Pish Commission 

Woods Hole, Massachusetts. 


Assistant Professor of Zoology , Hansard Unii^ersity, 

It is a well-known fact tliat many fishes are extremely sensitive to disturbances 
in the water such as are caused by splashing with an oar, stamping in a boat, or 
striking the side of an aquarium. When, for instance, tlw oi>a<j[ue wall of a fish 
tank containing young king-fish, sea-robins, or killi-fish is struck a vigorous blow 
with the fist, the fishes usually respond by giving a short, quick leap, and, if such 
blows are frequently repeated, surface fishes are often driven to the bottom and kept 
there. Notwithstanding the sensitiveness indicated by such reactions, most of these 
fishes appear to be unaffected by loud talking or other like noises originating in the 
air. Fishermen are familiar with these peculiarities and often Uike them into 
account in the practice of their art. 

Such facts as these are also usually accepted as evidence that fishes can hear 
(as an example, compare the statements made by W. C. Harris in Dean Sage's 
"Salmon and Trout," 1002, p. 311), but a simple experiment will show, I believe, 
that this assumpti(m is not necessarily correct. If one end of a wooden rod is 
vigorously tap[)ed while the other is beneath the level of the water a disturbance 
is produced that will call forth an obvious response from most fishes of moderate 
sensitiveness. Such a disturbance will likewise affect a human l>eing, for if one 
holds the head beneath the water the vibrations from the rod can be (^asily heard, 
and if the hand lx» placed in the water near the rod they can be distinctly felt. 

Since, as Miiller (1848, p. 1220) long ago pointed out, we can feel as well as hear 
these vibrations, it follows that such evidence as that already given can not be 
accepted as conclusive proof that fishes hear, for it is conceivable that their responses 
may be entirely through their sense of touch, i. e., dependent on their skins. More- 
over, fishes possess a special system of tegmentarj* sense organs, the lateral-line 
organs, which are completely absent from us, and it may be that these are in some 
way the recipient organs for the disturbances already descril>ed. When, therefore, 
a fish responds to water vibrations of the kind mentioned, we are not justified in 
concluding that it hears, for it may respond through the skin or the lateral-line 

organs and not through the ears. 




It may be reasonably asked at this point, What constitutes liearing? Ever}'- 
one will agree, I believe, that the sensation we get through the skin from a vibrating 
rod in water should not be called hearing, and what is true for us should hold for 
the lower vertebrates. Hearing in these animals may therefore l>e defined as tliat 
sensory activity resulting from a stimulation of the ear by material vibrations. This 
is in essential accord with the definition given by Kreidl (1895, p. 461) to the 
effect that hearing is that sensation which is mediated by the nerve that is homolo- 
gous with the auditory nerve of man. When, therefore, a fish responds to sound 
vibrations the question at once arises whether the stimulus is received by the skin, 
the lateral-line organs, or the ear. And until this question can be answered, at least 
so far as the ear is concerned, the (luery whether fishes hear or not must remain 
open. In dealing with this general subject I shall take up, first of all, the question 
whether fishes respond to sound vibrations through the eai's. 


Tnlrodurtory. — The internal eai's of fishes were descrilxid a.s early as Kilo by 
Casserius, and were studied in some detiiil in the following century by (4eoffroy, 
Scarpa, Comparetti, and Hunter. The attitude taken by many of thest^ early 
workers on the question of the ability of fishes to hear or not is well illustrat>ed by 
a quotation from Hunter (1782, p. 383), who at the conclusion of his paper on the 
organs of hearing in fishes made the following statement: 

As it is evident that fish possesH the organ of hearing, it lK»come8 nnnect^ssary to make or 
relate any exi)eriment made with live fish which only tends to i)rove this fac^t; hnt I will mention 
one exx)eruuent to shew that sound affects them much and is one of their guards, as it is in other 
animals. In the year 1762, when I was in Portugal, I olwerved in a nobleman's garden, near 
Lisbon, a small fish-jiond full of different kinds offish. Its bottom was level with the ground and 
was made by forming a bank all round. Tliere was a 8hrubl)ery close to it. Whilst I was lying 
on the bank, observing the fish swimming about, I desired a gentleman, who was with me. tx) take 
a loaded gun and go behind the shrubs and fire it. Tlie reason for going l>ehind the shrubs was 
that there might not he the least reflection of light. The instant the report was made the fish 
appeared to be all of one mind, for they vanished instantaneously into the mud at the bottom, 
raising, as it were, a cloud of mud. In about five minutes after they began to api)ear, till the whole 
came forth again. 

This passage show^s very clearly that in the opinion of Hunter the internal ears 
of fishes, like those of the higher vertebrates, are organs of hearing. Without further 
experimental evidence this view was accepted by Miiller (1848, p. 1238) in his well- 
known chapters on the physiology of the senses, and by many other emin<»nt authori- 
ties, such as Owen (18GG, pp. 342 and 346), Gunther (1880, p. 110), and Romanes 
(1892, p. 250). To these investigators the presence of the internal ears s<*emed, as 
it did to Hunter, sufficient ground for concluding that these aninmls could hear. 

Within recent years, however, this opinion has been call(;d in question, or even 
denied. Some of the grounds for tliis change of view may be stated as follows: 
Bateson (1800, p. 251), in some investigations on the sense organs and perception of 
fishes, observed that the report from the blasting of rocks caused conger's to draw 
back a few inches, flat-fishes (like the sole, plaice, and turl>ot) to bury themselves, 
and pouting to scatter momentarily in all directions; other fishes seemed to take 
no notice of the report. When the side of a tank containing pollock or soles was 
struck with a heavy stick, the fishes behaved as they did toward the report of the 


blasting. Pollock did not respond, however, lo the sound made bj^ rubbing a wet 
linger on the glasss window of an aquarium or to the noise made by striking a piece 
of gla^s under wat<ir with a stone, provided the means of producing the noise was 
not seen by the fishes. Bateson concluded that, while it may In? regarded as clear 
that fishes i>erceive the sound of sudden shocks and concussions when these are 
severe, they do not seem to hear the sounds of bodies moving in the water but not 
seen by them. 

Without knowledge of Bateson's observations, Kreidl (1805) carried out a series 
of experiments with the view of testing the powers of hearing in the gold-fish. Cur- 
assius auratus. This species was chosen iK^cause of the ease* with which it could be 
kept in the laboratory and, further, V)ecause it is one of those fishes that have long 
been reputed to come at the sound of a bell. Aft^n* an extended series of experi- 
ments, Kreidl (1805, j). 458) concluded that normal gold-fish never respond to sounds 
produced either in the air or in tlie water, though tliey do i-eact to the shock of a 
sudden blow given to the cover of tlie aquarium. Individuals rendered abnormally 
sensitive by strychnine gave no resinmse to the sound of a tuning-fork or a vibrating- 
rod even when these were in contai't witli the water, though the fishes resi>onded at 
once to such slight shocks as tapping the iuiuarium, etc., or even clapping the 
hands vigorously in the air. 

To test whether these response^^ were dependent ui>on the auditory nerves, 
Kreidl removed these nerves and the attached ear-sacs from a number of individ- 
uals, and, after poisoning them with strychnine, subjected them to stimulation by 
sound. In all cases they were found to respond precisely as the poisoned animals 
with ears did. Kreidl, therefore, concluded that gold-fishes do not hear by the 
so-called ear, but that they react to sound-waves by means of an esi)ecially developed 
cutaneous sense, or, to put it in other words, the gold-fish /e^fe sound but does not 
hear it (Kreidl, 1896, p. 581). 

After having reached this conclusion, Krei<ll was led to take up a si)ecific case 
of the response of fishes to the sound of a lx?ll, and an opportunity for doing this 
was found at the Benedictine monastery in Krems, Austria. Here the trout of a 
particular basin were said to come for food on the ringing of a lx»ll. Kreidl (1896, 
p. 583), however, found that they would assemble at sight of a pei-son and without 
the ringing of the l>ell. If they were not then fed, they soon dispersed and no 
amount of bell-ringing would induce them to return. If, however, a pebble or a 
small piece of bread was thrown into the water they immediately swam \igorously 
towanl the s|>ot where the disturbance had occurred. Moreover, if a person 
approached the ba>»in without being seen and rang the bell vigorously no response was 
observed. From these facts Kreidl (1806, p. 584) concluded that the assembling of 
the fishes was brought al)out through sight and the cutaneous sens<», and not through 
hearing, and that the conclusion reached with the gold-fish might be extended to 
other kinds of fishes. 

KreidFs conclusions were supj)orted by the observations of Lee (1808), who 
studied the reai^'tions of several species of fishes to such sounds as the human voice, 
the clapping of hands, and the striking of stones together in air and imder water. 
In all of his experiments Lee (1808, p. 137) obtained no evidence what<?ver of the 
existence of a sense of hearing, as the term is usually employe<l, although he found 
that the fishes were exceedingly sensitive to gross shocks, such as the jarring of their 

^ 48 



tank or coiieiissioiis upon its wallH. Lei% inoiH*over, called attiMitioii t'O the fa<*t that 
tlie papilla aciistica basilaris, which is the 8|>ecial or^an of hearing in the internal 
eara of the higher vertebrates, did not occur in the fishes. From the observatioiiB 
and experiments of Bateson and of Kreidl, and from his own work, Lee (181)8, p. 138) 
believed that the conclusion was justified beyond doubt that fishes do nof jK^ssess the 
power of hearing, in the sense in which the term is ordinarily used, and that the 
sole function of the ear in fishes is equilibration. 

The generalization to be drawn from the work just summarized, viz, that fishes 
do not hear, though they may respond to sound-waves by the skin, lias seemed to 
me not wholly in accord with certa^in well-known factw in the natural history of these 
animals. Among these facts may be mentioneil the undoubted ability on the pari, 
of some fishes to make sounds. If a fish has this power it might naturally l>e suj)- 
posed to hear the sounds it makes. Lee (181)8, p. 137) has (*alled attc*ntion to the 
small numlKjr of sound-proilucing fishes as evidence against, the vi<»w that fishes in 
general hear. But the fact that there an* such fishes has always apiK^aled to me 
in quit.e the reverse way and should, in my opinion, serve to indicate the siH»cies 
most worthy of attention in an}' investigation of the sense of liearing. It must l>e 
admitt>ed, however, that fishes may iHjssibly i)r<Hluce sounds that tlu»y themw^lves can 
not hear, but that other animals may hear and take warning from. Thus when small 
swell-fish, ChUomycterus scluvpfi^ are thrown into a t^iiik (*ontainiiig hungry scup, 
Stenotoiuu.H chrysopSy they are immediately set upon by the latt<»r. In defense* the 
swell-fishes inflate themselves with sea water till their tegmentary spines st^ind out 
rigidly, and at the same time they make a jwculiar sound by gritting the two front, 
teeth of the lower jaw against the inner surface of those* of the upiK?r jaw. It is not 
known tliat this sound is heaitl by the swell-fish, though it may Ik?. All that one can 
say with cert.4iinty is that the sound seems t4) Ik* directed against the f<M», for it. is 
made, so far as I know, only when the swell-fish is molested. (Tranting, however, 
that the swell-fish does not he^ir it« own sound, one would still Iw rash to conclude 
that this wjus an argument against the hearing of fishes, for the vast majority of 
animals towanl which the sound is directed are fishes themwdves, and tlu^se pre- 
sumably hear the sounds. 

Another good instance of the pro<luction of sound by a fish is found in tlie 
squeteague or weak-fish, Cytiosrion. re^jnlis. The grunting noise made l)y this fish is, 
however, pro<luced only by the males, and thisspecializ^ition is very difiitMiIt to under- 
stand unless one assumes an ability on the part of one or other sex to hear. Since 
the sounds made by both the sw^ell-fish and the squ<»t.<»agiie aiv in no sense shocks or 
concussions but resemble more closely, in rate of vibration and in intensily, su<*h 
sounds as might Iw* obtained from the ordinary a<*tion of an instruimMit like a tuning- 
fork of low pitch, it seems to me that they affonl evidence in favor of the s(»nse of 
hearing rather than the reverse. 

A second reason for questioning the generalization adv(wated ]»y Kividl, and 
by Lee, is the character of the observations upon which it is based. Both authors 
state that no positive evidence in favor of hearing could l>e obtained. But it must 
be borne in mind that in many animals known to possess a sense of hearing the 
auditory reflexes are perhaps the least conspicuous of any <'onnec»ted with the mon» 
important sense organs, and that consequently the most careful s<*rutiny of the 
movements of fishes must be made lief ore one can with certainty <le<»lare that hear- 
ing is absents. A perusal of the papera already summarizc^d led me to the conclusion 

1. side view ot nqiiarliini. nhnwlnit HOunillTiR apparatus at right-hand end and Kiupendnl kIbw rsK^ in whirh thi 

f. End Tlvw of Hqiurlum, nhnHtnR Hnindlng appaniliw. 

J. Ihiral viewol bmln ol Fnnilnl»i hrirroriaiii, dinn-led toahow piMlllOTUicil (he niola ol the flilh and »evunlli iitrvps (VI, 

the ninth and tenlh nenw (X). and thi> liilfmal eiir an inillpHItd t>- Its otolith. EnlarRpd alKiiit IwiiT. 
t. n<imi view ol head ot>VndrJ>w*rtnm-H(B».lo«hmrthi'r«rtnr where thp followiiut iiorvpBworLM'uI: The lilih hhAmvw 

eiKhth (Villi, and Ihc latrraHlne nerve (Xi. KnUntcd about twfec. 
i. aide view of >VmJH(««ArtrrtWi(M».i"howing the re([i«U¥rheiillhe\iWT»V\\n«iii;rte'«ttata.V ^,■J^^- W>1.t>\\N w^^*-'^''^- 




that something more might be attaiued in this direction, and I therefore resolved to 
give particular attention to the reactions of a few fishes with the view of ascertaining 
whether or not thev showed anv evidence of hearing. 

At the outset I thought it lx*st to experiment on some common sound-producing 
si>ecies, and for this purpose I did some preliminary work on the swell-fish {Chilo- 
mycierus schapji), the squeteague (Cynosrion rf^ijalis)^ and the sea robin {Prionoiiis 
caroUnu.s). To all of these, i)ractical object ions were found, and I was at last obliged 
to abandon them for (ishes that produce no sounds. Among these, three 8j)ecie8 
were found to be especially sensitive to slight vibrations — the king- fish (^fenf^rirrh^^s 
saxaiilis), and the two common s|>ecies of killi-lish (Fufuluhts mnjnlLs and F. hefero- 
clitii.s). Because of the great abundance of F. heterocJitns, the ease with which it 
could be operated upon, and its great hardiness, I cliose it for study, and the observa- 
tions recorded on the following pages, unless otherwise* stated, refer to this species. 

The ears in Funflnlus heteroclUus. — When a tank containing a numl)erof Fun- 
diilus helerocUtus is struck with the open hand so that the fish can not see the 
movement of striking, they respond to the vibrations by springing suddenly an inch 
or so through the water. The question to be considered is whether thest* vibrations 
stimulate the fishes tlirough the skin, the lateral-line organs, the ears, or some 
combination of these. If it could be shown that the eai-s were not stimulated bv 


the vibrations, it seems to me that we would have evidence pointing to the conclusion 
that the fishes did not hear. If on t lie other hand it could be demonstrated that the 
vil)rations did stimulate theeai*s, the evidence would l)e conclusive that the animals 
possessed the sense of hearing. To test these points considerable experimentation 
was necessarv. 

Much of the work that has Ihhmi carried out heretofore has l>een done with 
sound generated in air l)ut intended to affect fishes in water. That this method is 
extremely inefficient I found by trying the following exixM-iment. If a dinner liell 
is rung in the air l)y a person standing breast-deep in water, it will, of course, l>e 
heard easily by a second i)erson standing in a similar way a yard or two off. If, 
however, the second person i)Uts his head under the water during the ringing of the 
bell th(* sound seems to eease almost entirely and is not again heanl clearly by the 
diver till he emerges. In like manner a Ixdl rung or hit with a stone under water 
is heard, at best, very faintly by a jM^rson standing in the water unless his head is 
under the surf a<*e. In other words, the plane separating air and water is, under 
ordinary circumstances, an almost impenetrable one for most sounds, whether they 
are generat(»d on one side or the other of it, and man^ of the negative results 
obtained by previous invest igatoi*s on the sense of hearing in fishes may have bi^eu 
due not so much to the absence of hearing in the animals experimented ui>on as to 
their inaccessibilitv to the sound, or at least to sound of an inlensitv sufficient to 
stimulate. This difficulty has been recognized by Kreidl, and in devising apparatus 
I have profited by his exi>erience and used sound-producing appliances that were 
in direct contact with the water containing the fishes. 

The chief piece of apparatus that I used consisted of an ordinary marine 
aquarium (pi. i), Vi^. 1) with a slate base, two heavy glass sides, and originally two slate 
ends, one of which, however, I replaced l)y a pi(»ce of deal boanl fret* from knots, to 
serve as a sounding-l)oard. The inside dimensions of the aquarium were as follows: 
depth, 40 cm. {h* in.); breadth, 37 cm. (15 in.); and length, 87 cm. (35 in.). To the 
middle of one edge of the sounding-board a stout beam of wood was attached 

F.C.B. 190a-4 


SO that it stood out horizontally about 1 met^r (40 in.) in the plane of that end 
(fig. 2). From the free end of the beam a bass-viol strinj^ was stivtehed to the 
opposite side of the sounding-lx)ard. This string could be tightened by a bolt and 
nut at the free end of the ]>eam, and it was made to pass over a bridge placed neai* 
the middle of the sounding-lK)ai'd. The length of the string from the attached end 
on the sounding-board to the bridge was 25 em. (10 in.), and from the bridge* to 
the attachment near the free end of the beam 1.15 metei*s (45 in.). Thus the end 
of the aquarium might be regarded as something like a large one-stringed bass viol 
resting sidewise, with the sounding-board for a lH>dy and the lK*am for a neck. 

When the string was tightened and plucked or bowed a good tone was obtained, 
which was transmitted directly through the sounding-board to the wat<*r within the 
aquarium. On keying the string up to a good clear tone, I found by writing off its 
vibrations on a revolving cylinder that it produced on an average 40 per second, 
and I retained this pitch by frequent adjustment for the exi)eriments that I subse- 
quently performed. I was led to adopt this low tone since most of the noises that I 
have heard fishes make wen^ in the nature of low-pitched grunts. 

Each time the string was plucked the note began wuth maximum intensity and 
then gradually died away. It was, consequently, impossible to get any very signifi- 
cant record of the intensity, but I endeavored to use the apparatus in a uniform way 
by drawing the string out a constant distance from its position of rest each time 
I plucked it. The distance usually employed was about 1.5 cm. (0.0 in.). The 
amount of weight required, when liung at the middle of the longcir segment of the 
string, to depress it 1.5 cm. was found to be about 2.15 kilograms (4.75 i)ounds), 
so that each time the string was liberated on being plucked in the usual way, it 
moved forward with an initial force equal to the pull of 2.15 kilograms, a rough 
measure of the maximum int<»nsity of the sound produced. 

The fishes to be experimented upon were not allowed to swim unrestrict^^d in 
the aquarium, but they were placed in a small cage (fig. 1) susi)endtHl from a cord 
attached at its ends to the walls of the room. Thus the support for the fish cage 
was entirely independent of the walls of the aquarium and any vibration that reached 
the fishes must have done so almost entirely through the water. The cage could be 
moveil in a horizontal direction back and forth on the cord, and thus the fish could 
be placed at any desired distance from the sounding-board up to 75 cm. (»J()in.). 
The inside measurements of the cage were as follows : Height, 10 cm. (4 in.); length, 
20 cm. (8 in.); and brea<lth, al)out 10 cm. (4 in.). The bottom of the cage was wood, 
padded on the inside with cotton wool covered with cloth to provide a deadened 
surface on which the fishes might i*est. The top and three sides wen* glass; the 
fourth side was made of coarse netting to retain the fish but to interfere as little as 
possible with the entrance of sound, and this side was always <lirected toward tin* 
sounding-board. As the fishes averaged about 7 cm. (2J in.) in length, the cage 
gave them ample room for moving about. 

My plan was to introduce fishes in various conditions into the cage, and, after 
they had become accustomed to their surroundings, to subject them to stimulation 
by sound and observe their reactions. I found it desirable to exiH»riment with three 
clasFCs of fishes; first, normal ones for a basis of comparison; secondly, fishes from 
which the ears had been removed; finally, fishes in which the general integument 
had been rendered insensitive, but in which the eai-s were intact. The methods of 


obtaiuiug fishes in tliese conditions and the resi)onses that they showed will be 
described for each class of fishes. 

Xoniial fislie^. — When a normal fish is first lilx* rated in tlie caj?e it swims vigor- 
ously about for a few moments, after which it may, sooner or later, come to rest on 
or near the bottom. The animals are exti-emely quick-sighted, and, if aft^r they 
have come to rest the observer makes any sudden movements near the a^iuarium, 
they are very likely to begin active swimming anew. It is, therefore, extremely 
necessary to work in such a way that all movements, and particularly quick ones, 
are made out of sight of the fish. When the fish is resting on the lK)ttom of the 
cage, two sets of motions will usually be observed: first, the respiratory movements 
of the operculum; and secondly, the alternate vibratory movements of the i)ectoral 
fins. The oi)ercular movements, as might 1h» exi)ected, always continue, but the 
movements of the pectoral fins, which seem to be connected also with the respiratory 
function, often cease entirely. 

When a fish has become quiet, except for the respiratory movements, the vibra- 
tions from the string nmy call forth any of four kinds of resi)onses. The first of 
these is the vibratory movement of the pectoral fins, either a few slight beats, if the 
fins were previously at rest, or an increased rate or extent of swing if they were 
previously in motion. The vibration of the string at the intensity ordinarily 
employed almost invariably called forth this reaction; thus, in ten observations 
t^ken from each of ten fishes at a distance of about 25 cm. (10 in.) from the sounding- 
board there were iMj pectoral-fin responses and 4 failures. Since this response is so 
readily observed, it has affonled one of the most satisfactory criteria of stimulation. 

The second form of response is a change in the rate of the respiratory movements. 
In a ([uiescent fish measuring 8 cm. (3.2 in.) in length the resi)iratory rate was 114 
per minute. On stimulating by sound this rate rose suddenly to 138 per minute for 
some ten or a dozen movements and then fell rapidly to about the former numl^er. 
This is x)robably a very usual form of response, i^erhaps <iuite as much so as the 
movement of the pectoral fins, but the shortness of its duration and its inconspicu- 
ousness make it less satisfactory as an indication of stimulation than that afforded 
by the i)ectoral fins. If the sound from the string is of considerable intensity, the 
third form of response may appear, a slight motion of the caudal fin, l)eginning 
usually at the doi-siil edge and proceeding as a wave ventrally. Finally, with strong 
stimulation, the fish may make a short but <iuick spring forward. 

All these ivactions have been obtained from fishes even at 75 cm. (30 inches) 
from the sounding-lx)aiil, although the springing movements are more frequently 
observed when the animals are not so far from the source of sound. One verv inter- 


esting fact alx)ut these reactions is that they can not be repeated rapidly for even a 
short period. A fish that responds to the first stimulus by a spring, may react to the 
second or to the third only by moving the pectoral fins, and to the fourth in no 
observable way. It is only when a considerable period of rest intervenes that the 
reactions may be repeatedly obtained; and I have found that the minimum period 
of rest is not far from one minute, though, even then, reactions may sometimes fail 
to appear. 

Earless fishes. — The ix^moval of the ears from a fish is a serious operation, but it 
is one which, after a little practice, may be accomplished with success and from 
which the fishes generally recover. These animals are easily etherized by putting 


them in sea water containing enough ether to give it a strong odor. On being trans- 
ferred to pure sea water they quickly recover, and an individual that I etherized six 
times in the course of one afternoon finally recovered without showing any ill effects. 
The first method I used in operating on the etherized fishes was to oi>en the cranium 
in the region of the ears and, after cutting the auditory nerves, to remove those 
nerves and the attached ear-sacs. These i)art8 were easily identified from the fact 
that the auditory nerve emerges from the medulla almost exactly ventral to the cleft 
between that organ and the optic lobe, and the ear-sac, which is only partly sur- 
rounded by cartilage, lies in the cranial cavity only slightly peripheral to the point 
where the nerve leaves the medulla (pi. 0, fig. 3). After the operation the fishes 
were returned to pure sea water and, notwithstanding the exposure of the brain, a 
considerable number recovered and survived. One of these I kept for more than six 
weeks, and, though its swimming was characteristically irregular, it was alert and 
active and, except for a brief intervening period, it fed normally. 

From the operation just described about one fish in t^n recovere<l. This pro- 
portion was greatly increased by a second form of oi)eration in which the auditory 
nerve was cut without oi)ening the cranium (fig. 4). After a little practice I found 
that this could be done with great certainty and about eight out of ten fishes usually 
recovered. All fishes that had been operated on were kept at least twenty-four 
hours before they were subjected to experimentation. 

Fishes in which the auditory nerves have been cut have very characteristic 
reactions. When resting or when swimming slowly they behave for the most part 
as normal fishes do, and, in fact, are often undistinguishable from individuals upon 
which no operation has been performed. When, however, they are stimulated to 
rapid locomotion, they swim either in irregular spirals, the same individual revolving 
sometimes to the right and sometimes to the left, or they turn over and over in 
irregular circles without a(*complishing much real progression. This loss of 
orientation on attempting rapid locomotion has for some time been recognized jis 
indicative of one of the chief functions of the ears in fishes — i. e., equilibration. 
It is probable that in resting or in swimming slowly the fish dei)ends uiK)n the eye 
for orientation, but in quick movements the ears come more into play, and hence 
after their loss quick movements are accompanieil with lack of orientation, llie 
forced movements thus observed may be t.iiken with perfect cert^iinty, so far as my 
exi)erience goes, as evidence of tlu? successful out<.»ome of an attempt to cut the 
auditory nerves, for in the few cjises where thesis movements failed to appear, 
subsequent dissection showed that thci nerves had not l)een cut, and in all inst^inces 
where the movements were observed and the animals afterwards disse<»t<Hi, the 
nerves were found severed. 

A second feature of interest that generally character iz(^d fishers with severed 
auditory nerves was the color that they finally assumed. I'^nder ordinary circum- 
stances the color of this species is a light greenish-gray. When etherized the fishes 
become very dark, with a mottling of blue-green on the sides and belly. After 
recovery from cutting the auditory nerves, the dark coloration disappears and the 
fish assumes a tint even i>aler than that of a normal individual. This tint is retained 
throughout life. Etherizing prebably influences the chroinatophores of the skin 
directly, but cutting the auditory nerves introduces changes that are i)n)bably 
dependent upon the nervous control of the chromatophores. 


When the earless fishes were tested in the sounding apparatus, they yielded 
very interesting results. Unlike the gold-fishes experimented on by Kreidl (1805), 
they differed markedly from normal fish. In an extended series of observations on 
over 20 fishes I never once observed with eertainty the springing reflex as a result 
of sounding the l)ass-viol string. The fishes were usually very aetive, and I was 
never abh* to ascertain with eertainty whether they showed a change in the respira- 
tory rate on stimulation. The ix»etoral-fin movements, however, were observed with 
mueli eertainty. On 10 earless fishes I succeeded in getting 10 observations each 
to sound stimuli at al>out 25 cm. (10 inches) from the sounding-board. The total 
result was that in 82 observations there were no reactions and in the remaining 18 
the reactions were at best slight ones. As the fishes often moved the pectoral fins 
without apparent cause, some of the 18 reactions may have been accidental coinci- 
dences, but others were so precise and typical that I am convinced they were due to 
stimulation. Earless fishes, therefore, differ from normal ones in that their pectoral- 
fin resix)nses to vibrations from the bass-viol string are enormously reduce<l, though 
not entirely obliterated. 

Fishes with insensitive skins. — For reasons already given it isimp)erative, l)efore 
drawing conclusions from the condition of earless fishes, to examine the evidence 
afforded by those whose general sui-face has l)een rendered insensitive. In this way 
it is i)ossible to ascertain what part the integument plays in the reception of sound 
vibrations. I had hoi)ed that the integument of Fnndulus heteroditus could be 
rendered insensitive by immei*sing the fish for a short time in a solution of cocaine, 
but all attempts in this direction proved failures, since the drug acted much moi'e 
vigorously as a poison than as an anaesthetic, and I was finally obliged to abandon 
this method altogether and resort to nerve-cutting. 

The following operation performed on etherized fishes insures an almost complete 
insensibility of the surface. The fifth and seventh cranial nerves can l)e cut just 
posterior to the eyeball (pi. 0, figs. ^5 and 4), the lateral-line branch of the tenth nerve 
can next be cut at tlu* posterior edge of th(^ pectoral girdle (fig. 5), and finally the 
spinal cord can be sevi^red at the fourth or fifth vertebra. Severe Jis this operation 
is, almost all fish<\s recover from it and n^spire and feed normally, though they 
seldom live Ix^vond two w(»eks after the operation. 

Fish<»s that have recovennl from this oi>eration show certain well-marked charac- 
teristics. The integument, particularly that of the dorsal surface, is unusually 
dark, as a r<\sult of the expanded condition of the chromatophores. The fish's 
mouth is gaping and motionless in consequence of the motor portion of the fifth 
nerve having Ix^en cut. This condition, however, does not interfere with respiration 
or with the sucking in of pieces of food, an act which the fish performs with avidity. 
Since in cutting the fifth and seventh nerves, the three small ner\'es to the muscles 
of the eyeballs, the third, fourth, and sixth, must also be cut, the eyes are motion- 
less And usually i)rotru(le somewhat. Finally, as a result of cutting the spinal cord, 
the whole trunk of the aninml is, as a rule, passive and is drawn after the head, the 
swimming being perform(»d by tlu* piH'toral fins. Since the greater part of the conl 
is intact, a mor<» or h»ss vigorous stimulus applied to the trunk is followed by move- 
ments in the doi-sal, anal, and caudal fins, or even by a locomotor response of the 
whole trunk, but such movements are made only after special stimulation, and the 
trunk is onlinarily carried passively, like a paralyzed appendage. As a result of 


having so little of the normal locomotor apparatus intact, the fishes often swim ven- 
tral side up, for the action of the pectoral fins is not always sufficient to overcome 
the physical effects of the si)ecific gravity of the fish's bo<ly. 

Fishes that have recovered from the operation just descrilx^l liave intact the 
ears, tlie central nervous organs from the anterior end of the l)rain caudad to the 
fourth or fifth vertebra, and the sensory and motor apparatus for the region of 
the gills and the pectoral fins. Excepting in tliese two ratlier restricted regions, the 
whole integument is insensitive, at h^ast so far as its <*apacity to originate impulses 
to movements in the gills or pectoral fins is concerned. Such fish, then^fore, are in 
a condition to receive stimuli through the ears and to respond by ivspiratory or 
pectoral-fin movements. 

The reactions that these fishes showed to the sound apparatus were surprisingly 
clear and decisive. From the nature of the o|)eration one would not expe(»t them to 
l)e able to give the sudden spring that the normal fishes often sliowed, and, as a 
matter of fact, such responses were never observed. Were the skin of tlie trunk 
sensitive, it is conceivable tliat the (»audal-fin reaction might o<?cur, for the conl, 
though severed from the rest of the central nervous organs, was in itself inta<*t. 
Caudal-fin reactions were, however, also never observed. The resi>iratory reactions 
and the pectoral-fin responses occurred with great i*egularity. When tlie bass-viol 
string was made to vibrate, the respiratory rate increased for a very brief period. 
Ii^ a fish 7 cm. (-J inches) long the rate previous to stimulation wjus 120 per minute; 
immediately after stimulation it wjis 156. The reactions of the pectoral fin were also 
well marked. In ten observations on each of ten animals at a distance of about 25 
cm. (10 inches) from the sounding-boai*d the pectoral-fin responses occurnMl 'J4 times 
in the total hundred. This is in close agreement with the normal fishes and in strong 
contrast with the earless ones. So far, then, as I'eactions to the vibrating chord are 
concerned, these fishes show the essential characteristics of normal individuals. 

Discussion of the results of the experiment's. — It is clear from the experinu»nts 
descrilwd in the pit»ceding sections that fishes wliose eai*s were rendered functionl(\ss, 
but whose skins were normally sensitive, reacted only slightly to the stimulus from 
the sound-produidng apparatus, whereas those with insensitive skins but functional 
ears responded to this stimulus, as far as their conditions would permit, almost exa<*t ly 
as normal fishes did. It might be assumed that the failure to respond on tlu» part of 
earless fislies wjis due not to the loss of the ear, but to the shock of the operation 
they had un<lergone. This, however, does not seem to l>e the, for, after the 
fishes had recovered from tlie immediate effects of the operation, th(\v were* active, 
fed well, and sometimes lived many weeks. Moreover, if the op(*ration w(»re as 
severe as is implied in the above assumption, one might exi)ect sonu* indications of 
this in fishes in which onlv one auditorv nerve had lx>en cut. As a matter of fact, 
immediately afteir this operation fishes with only one ear intact did swim irr(»gularly, 
but in from six to eight houi-s this tendency disappeared entirely, and tin* fish in its 
quickness, precision, and noimality of response ]>ecame, so far as my observations 
went, absolutely indistinguishable from a normal individual. Further, fishes with 
the fifth, seventh, and lateral-line nerves and spinal cord cut have without doubt 
suffered a more severe shock than those that have had only the eighth nerve cut, 
and yet the pectoral-fin reactions of the fonner were essentially normal. It therefore 
seems to me that the great reduction in the number of pectoral-fin reactions of 


earless fishes is due to the loss of the ear as a sense organ and not to seeondary 
complications accompanying the operation. 

Although some of the observations recorded on the preceding pages make it 
certain that in these fishes the eai-s are stimulated by disturbances such as those 
st»t up in the water by the sounding apparatus, it may still fairly be asked whether 
these disturbances are in the nature of sounds. When the bass-viol string attached 
to the a<iuarium wjis plucked, a series of sound waves of diminishing intensities 
was delivered to the water. To ascertain something of the nature of this sound I 
immersed my head in the water of the aquarium and had an assistant pluck the 
string in the usual way. The sound that I thus heard was, so far as I could judge, 
of nearly the same i>itch as that which the string gave to the air and of only slightlj' 
greater intensity. This sound certainly reache<l the fishes. 

The sounding apparatus, however, did more than give rise to this sound. When 
the stringwas plucked two things besides the production of sound certainly happened: 
First, the whole aquarium, including its supi)orting table, ti'embleil slightly, and, 
probably as a consequence of this, ripples started from the ends and sides of the 
aquarium and proceeded toward the center. These ripples, though chiefly surface 
effects, indicated a wave motion that penetrated the water to some extent, and that 
was doubtless the cause of the very slight swaying movement of the fish cage occa- 
sionally noticed after the string had tn^en vigorously plucked. Moreover, a distinct 
tremor could l)e felt in the water when the hand was held 5 to 8 cm. (2 to 3 inches) 
from the sounding-board and the stringwas plucked. The question naturally arose 
wheth<»r the fishes did not resiK)nd to the movement of the aquarium as a whole or 
to the wave movement indicated l)y the ripples rather than to the true sound waves. 

To answer this question, at least so far as the ripple movement was concerned, 
I was led to study tlie reaction time of the fishes. Unfortunately circumstances 
prevented me from reducing this to a very accurate process; but, by listening to the 
beat of a chronometer and at the same time watching the fish, I am confident that 
the fin reactions occurred in less tlian 0.2 s<*cond after the string had been plucked. 
The sound waves and ripples mentioned above traveled from the sounding-board 
toward the fish at very different rates. The sound waves must have passed over 
the 25 cm. of water betwcHMi th<» sounding-board and the fish almost instantly. The 
surface ripple traveled much less rapidly and its rate couhl be easily measured. 
This proved to be a meter {'M^i inches) in 4.8 seconds; hence, to travei-se 25 cm. (10 
inches) the ripple re<iuired alK)ut 1.2 seconds. Since the fislies responded in less 
than 0.2 .second, they must have reacted to something other than the disturbance 
indicated by the ripples. 

Having eliminated the ripples as the initial stimulus for the fishes, it remained 
to l)e shown whether this stimulus was the movement of the whole aquarium or the 
sound waves proper. I succee<led in doing this by substituting an electric tuning- 
fork for the bjiss-viol string. Th<» tuning-fork was placed so that its base was within 
about a millimeter (.^K iiirh) of the sounding-board. The iron frame holding the fork 
nested on supports nuide of rubl>er bottle-st()pj)ers. These flexible supports alio wtnl 
the fork to be moved enough to bring its basi* into contact with the sounding-board 
without moving the supports owv the surface on which they rested. As this could 
be done without any initial jar, it was possible to communicate to the water in the 
aquarium a sound of uniform intensity an<l pitch without moving the aquarium as 


a whole and also without producing any ripple. The fork, moreover, produced a tone 
much purer than that obtained from the string. It had a pitch of 128 vibrations per 

Earl(»ss fishes, when subjected to sound waves from the tuning-fork, showeil 
notliing tliat I could identify as a reaction. Normal fishes and fishes with normal 
ears but insensitive skins very usually reacted by pectoral -fin moveuKMits. The 
occasional failure to respond was attributed by me to the faintness of the vibrations, 
for the most intense sound obtained from the fork wivs much than that pnKluced 
by the bass-viol string. That the fishes, however, always did rem^t, even to this 
relatively faint tone, was pointed out to me by ni}' friend Dr. F. S. Lee, who while 
watching one of the experiments thought he detected an increase in the respinitory 
rate even when no pectoral-fin reac^tion occurred. Subsequent study showed this to 
he entirely correct, for, irrespective of pectoral-fin responses, at each sounding from 
the tuning-fork an increase of the respiratory rate did take place for a very short 
period. There is, then, no question but that these fishes I'cspond to sound waves, 
and, since this response is through the ear, I conclude that Fundulns heferorlitus 
may be said to hear. Since I never succeeded in getting reactions of any kind to 
t)ie tuning-fork from earless fishes with skins and lateral lines inta(»t, I have no nnison 
for believing tliat these parts are stimulated by true sound wavers, and I attribute) 
the responses that earless fishes occasionally sliowed to the vibrating bass-viol string 
not to the action of its sound waves on the skin or the lateral-lin<» organs, but, as 
will be shown later, to the influence of the accompanying movement of the whole 
aquarium and it^s contained wati^r on these parts. 

Although the experiments already described remove every reasonable doubt 
from my mind as to the ability of these fishes to hear, the object ion may still lx» 
raised that the conditions under which they were carried out wei'e so artificial that 
they may he said to have almost no bearing on the ordinary habits of FunduJuSs and 
it must he admitt4?d that the relatively small volume of wat«»r in the aquarium and 
the character of its walls as reflecting 'surfaces for sound, may possibly have 
introduced factors to which the fishes, in their natural surroundings, weiv not 
accustomed. To ascertain how much weight should be given to this objection the 
following experiment was tried. The sounding apparatus, consisting of the sounding- 
board and the bass-viol string, was taken from the aquarium and set up in the open 
water of the outer pool at the Fish Commission wharf. The fish cage was hung at a 
distance of 50 centimeters (20 inches) from the sounding-board and toward the center 
of the pool, which is alwut 100 feet wide. The sound, therefore, was as unrestricted 
as that which naturally reaches these fishes. On exixM'imeiiting with normal fishes, 
fishes without ears, and those with insensitive skins, results were obtained ess(Mitially 
like those observed in the ac[uarium, and I therefore concluded that the restriction 
of the water in the aquarium played no essential part in the n\sults obtained from 
that apparatus. Then* is, thus, good reason to Indieve that Fundulus heffrorlifiLs 
not only hears, but that for it hearing is a nornml process. 

Having determined that h(*aring was one of the normal functions of the ears in 
Fundulu^y I had hojMHl to l)e able to ascertain by experiment th(^ parti<nilar part of 
the ear, if such there l)e, that was concerned with this s(Mis(». The internal ear in 
FumlulU'S heferorlitus \H like that in most, teleosts. It consists of the usual three 
semicircular canals an<i a large 8ac<Milus, at whose posterior end a well-develoiHKl 
lagena is present. The sacculus is a thin-walled chamlH»r, vcM'tically flatten(»d and 


containinf]^ a thin, flat otolith of considerable size. Sometimes this otolith is repre- 
sented b}' two pieces — a small one at the anterior end of the sacculus and a much 
larger one occupying the more central part of this chamber. The lagena, which is 
well separated from the sacculus, also contains an otolith. On the median face of 
the sacculus is an extensive macula acustica sacculi, formed by the termination of 
the major part of the eighth nerve. There is also a well-develojie*! papilla acustica 
lagenje, as wvll as the usual three cristfe acustica* ampullarum. I am unable to 
state whether other sensory patches, such as the macula acustica neglecta, occur 
here or not. 

Having made a preliminarj' study of the anatomy of the internal ear, I had hoped 
to Ik* able to cut in different individuals diffei'ent branches of the eighth nerve, and, 
by further expi^ri mentation on fishes thus prepared, to determine the functions of 
the s(*veral sense organs of the internal ear. After numerous unsucce.ssful attempts 
I was at last obliged to abandon this plan because of the small size of the branches 
and their somewhat intricate relations, and I am, therefore, in possession of no 
observations that show which part or parts of the internal ear are concerned in 


Infroducfonj, — The lateral-line canals were regarded by most of the earlier 
investigatoi-s as glands for the production of the mucus so characteristic of the skins 
of fishes. AlK)ut the middle of the last century Leydig (1850, p. 171) discovered the 
numerous sense organs contained in these canals, and declared that the whole system 
represented a sensory apparatus peculiar to fishes. Subsequently Leydig (1808, 
p. 2) expressed the opinion that these organs implied the possession of a sixth sense, 
one in addition to the five usually attributed to vertebrates, though he admitted that 
this sense was probably closely related to touch. Two years later the lateral-line 
organs were investigated by Schulze (1870), who demonstrated that true lateral-line 
organs were found only in the water-inhabiting vertebrates. From a study of their 
struct ui*e Schulze (1870, p. 80) was led to the lK*lief that they were stimulated by 
the mass movement of the water, as wiien a current passes over the surface of a 
fish or when the fish swims through the wat<*r. lie further believeil that they were 
stimulated by sound waves whose length was greater than that of waves to which 
the ear was adapted. In this respect they were organs somewhat intermediate in 
character between those of touch and of hearing. These opinions w^ere opposed by 
Merkel (1880, p. 54), who j^ointed out the inaccessibility of the organs to moving 
water in many cases, and who regarded them merely as organs of touch. The 
opposite extreme was taken by P. and F. Sarasin (1887-1890, p. 54), who designated 
them acces.sory ears, a view suggested some years previously by Emery (1880, p. 48). 

The opinions thus far given were base<l for the most part on an int<3rpretation 
of the anatomy of the lat<*ral-line organs, and not upon any positive experimental 
evidence as to the function of these parts. Fuchs (1895, p. 467) seems to have l)een 
the first to attempt work in this direction. His experiments were made chiefly on 
the torpe<lo, a fish in which, in addition to the lateral line proper, two other sets or 
organs, the vesicles of Savi and the ampulla? of Lorenzini, may be regarded as parts 
of the lateral-line system. In an active torpedo Fuchs cut the nerves connected with 
these two special sets of organs without, however, being able to detect any significant 
change in the subsequent movements of the fish. lie then exposed the nerve inner- 
vating the vesicles of Savi, and having placed it in connection with the appropriate 


electrical apparatus, he found that on pressing lightly upon the vesicles a negative 
variation in the current from the* nerve could be detected. As this negative varia- 
tion is evidence of the momentarily active condition of the nerve, it follows that 
pressure differences may be assumed to be a means of stimulating the vesicles of 
Savi. No such results were obtained from the nerves distributed to the ampullae 
of Lorenzini, but the nerves from the unnuxlified lateral-line organs in Raja davata 
and R, a^terlas showed negative variations when their terminal organs were sub- 
jected to pressure. Dilute acids and clianges of temperatui^e were not stimuli for 
any of the t(»rminal organs tested, an<i Fuchs (18!)5, p. 474) concluded that pressure 
was the normal stimulus in the skate for the lateral-line organs, and in the torpedo 
for the vesicles of Savi, but not for the ampulla? of Lorenzini. 

Apparently without knowledge of the work done ]>y Fuchs, Richard (181)(), p. 
L'U) performed some experiments on the gold-fish. These consisted in the removal 
of the scales from the lateral line and the destruction of the sense organs under 
these scales by cautc^rizing with heat, silver nitrate, or potASsic hydrate. Aft-er this 
operation some of the fishes were unable to keep below the surface of the water, and 
though they soon died, Richard (1800, p. 133) l>elieved that he had evidence enough 
to show that the lateral-line organs were connected with the production of gas in 
the hydrostatic apparatus. 

Richard's conclusions were called in question by Bonnier (189(5, p. 017), who 
pointed out the seventy of the operations employed by the former and intimated 
that the results were more i)robably dependent upon the excessive amount of tissue 
removed than upon the destruction of the lateral line. Bonnier (180(), p. 018) 
further recorded experiments of his own in which the lateral-line organs were 
destroyed by electro-cauterj'. Fishes thus operated upon showed two (*liaracter- 
istics — they could easily be approached by the hand and even seized, and they 
failed to orient themselves in reference to disturbances caused by Ixxlies thrown 
into the water. Bonnier concluded from his experiments that the lateral line, in 
addition to other functions, had to do with the orientation of fishes in reference to 
centers in the water from which shock-like vibrations might proceed. 

Lee (1808, p. 130), whose experimental methods were mucli the Siime a,s those 
used by Bonnier, obtiiineil some significant results, particularly with the t4)ad-fish, 
BairacJms tau. When thypectoral and p(»lvic tins of this lish were removed, so tliat 
the animal might be saicl^o l)e without it^s usual mechanical support, and the lateral- 
line organs were destrowid by thermo-cautery, the animal would lie quietly for some 
time, either on its side^r back, and acted as though it had lost its "sense of equili- 
bration." That its condition was not due to excessive injury was seen from the fact 
that a fmless fish in which an equal amount of skin had been cauterized, but in 
which the lateral-line organs were int4U*t, showed no lack of equilibration, and in its 
general behavior closely resembled a normal fish. Moreover, stimulation of the 
central end of the lateral-line nerve resulted in perfectly coordinated fin movements, 
and Lee (1808, p. 144) therefore concluded that the organs of the lateral line are 
equilibrating organs. How these are stimulat.ed Lee does not attempt to decide, 
though he suggests (1808, p. 143) that pressure changes in the surrounding medium 
may be the means of stimulation. 

From this brief historical resume it must be evident that there is still very little 
unity of opinion as to the functions of the lateral-line organs. 


TJielaferalline in FinifhihisheterocJitus. — The lateral-line system in i^. heterocUtus 
presents a condition typical for teleosts. Its sense organs are contained in canals 
that opi'n by pores on the surface of tho skin. A lateral-line canal as indicated by 
its pores (pi. 0, fig. 5) extends along the side of the trunk from near the tail forward 
to the head. Here tlie arrangement of the pores (figs. 4, 5) gives evidence of a man- 
dibuhir, a sul)orbital, a supraorbital, and an occipital branch. By cutting the fifth 
and the seventh nerves behind tlie eye (fig. 4), and the lateral-line nerve near the 
pectoral girdle {^\\::. 5), the innervation of this whole system, except a small tract 
al)ove tlie gills, can l>e Hindered inoperative; the sense organs in the small tract can 
be easily excised. Fishes that have undergone this oix*ration recover almost inva- 
riably and in a very sliort time*; the integument of their heads is insensitive owing 
to the necessity of cutting tin* fifth as well as the seventh nerves; but that of 
their trunks, which is of course innervat<Ml from spinal sources, retains its normal 
sensitiveness, except so far as the lateral-line organs are concerned. In seeking 
for evidence as to the function of the lateral-line organs, I compared carefully the 
reactions of normal fishes with those in which the nerves of the lateral-line organs 
had been cut. 

When a normal fish is lil>erated in an aquarium, it swims at once to the bottom. 
Here it may move about excitedly for some minutes, after which it usually begins to 
make upward excursions. At fii'st it will swim only part way to the top, returning 
each time quickly to the bottom. Eventually it may make several quick excursions 
to the upper surface of the water, and ultimately may remain there playing about 
close to the top. If now any disturbance is made the fish will again swim at once 
to the bottom, and only after some time will it return to the top, in the same cautious 
wav as lK»fore. Almost anv disturbance seems to drive the fish to the lx)ttom — a 
flash of light on the water, a quick but noiseless movement of the observer, or an 
unseen blow on the a([uarium, conditions all of which suggest that the movements 
of the fish are of a protective nature. 

To one form of disturbance the fishes were particularly sensitive, and this was 
the slight movement of the whole a<iuarium that occurred whenever the bass-viol 
string was jducked. This movement could be pro<luced without the accompanying 
sound by giving a slight vibratory motion to the beam attached to the sounding- 
lH)ard on the aquarium {^\::. 2). It was remarkable how accurately the fishes 
respond(Ml to this stimulus. If the fish was playing at the top of the water, the 
slightest movement of the aquarium as a whole would cause it to descend imme<li- 
ately to the bottom; if it was on its upward course, it could be checked and made to 
descend at any point; and if it was near the bottom, it could l>e kept there as long 
as the movement continual. In all of the several hundred trials of this kind that 
I made, I never found a normal fish that would remain high in the water or swim 
upward while* such movements were being imparted to the aquarium. WTienever 
the fish was above the lx)ttom, the resp<3nse was an instantaneous downward course. 

With fishos in which the nerves to the lateral-line organs had l)een cut, the 
reactions were totally difTerent. Such fishes, when left to themselves in an aqua- 
rium, wen^ scarcely distinguishable from normal ones. As with the toad-fishes 
observed by Lee (18118, p. 140), the loss of the lateral-line organs seemed to interfere 
in no essential respects with the movements of the animals; they were active and 
quick, returned at once to a normal position when displaced, and oriented with 


accuracy, so far as I could see, in that they at once swam away from such centers 
of disturbance as come from droppinj^ a stone in the water. In this hist partic'uhir 
they were very unlike the fishes reported on by Bonnier (1896, i>. 018). In one 
important point they <liffered absolutely from the normal fishes; they would swim 
upward and remain near the top during even a considerable agitation of the whole 
aquarium, though they would dart downward at any su<lden movement on the part 
of the observer. Hence these fishes must have lost their capacity to be stimuhited 
by the mass movement of the water, jind since this defect was observed only after 
the lateral-line organs had been rendered inoperative, I concluded that the normal 
stimulus for these organs was a yery slight mass movement of the surrounding 
water. Since such movements always accompanied the sound produced by the bass- 
viol string, it follows that the disturbances set up by this string must have acted as 
a stimulus for the latin*al-line organs as well as for the eara, and it is therefore not 
surprising that earless fislies sometimes reacted when the string was plucked. 

If the lateral-line organs are stimulated by a slight mass movement of the water, 
it occurred to me that I ought to be able to separate, in a mixed school of fishes, those 
with lateral-line organs intact from those in which the nerves to the organs had 
been cut, by simply imparting a slight mass movement to the water. Under such 
conditions the normal fishes ought to swim to the bottom, leaving the defective ones 
above; but cm trying the (experiment I found that the fishes \vi}w so accustonunl 
to form a school that when the normal ones started for the bottom th<» others <lid 
the same, and I was entirely unable by this means to separate the normal from 
the defective individuals. lUit I finally succeeded in doing this by modifying the 
experiment, in that I used only two indivi<luals, one normal and one defective, and 
agitated the aquarium only when they were widely separated. The result was very 
decisive in that the normal one invariably t/ook the initiative in descending, and in 
fact was often not followed by its defective companion. 

Having found the conditions under w^hich th(^ lateral-line organs were stimu- 
lated, it is natural to inquire as to the exact nature of the stimulus. Oixlinarily the 
fishes were induced to react by making the whole aquarium swing at about ten 
vibrations i>er second; but a like reaction was obtained from the normal fishes when 
a single swing, or what was as near as possible a single swing, was given to the 
aquarium. The stimulus therefoi*e is not n<?cessarily of a vibratory kind, but c<m- 
sists in a slight movement of the bcwly of water a.s a whole. It might be supposed 
that since the fish was suspended in the water, the motion of th(» ac^uarium *is a 
whole could have no influence on it. Hut it must be remembered that tlu* fish was 
somewhat heavier than the water, and that each time the aquarium was moved the 
fish, from its inertia, must have lagged a little behind or, once set in motion, moved 
a little ahead, and it is this slight difference in the rate of movement of the fish and 
of the adjacent water that, in my opinion, induces stimulation. I am not prepared 
to say how this affects the sense organs in the lateral-line canals; but it is not 
impossible, as Schulze (1870, p. 85) suggest<»d, that slight currents are thei-eby si»t 
up that move and thus stimulate the bristle cells of the lateral-line organs. 

The extreme sensitiveness of animals to slight motions of this kind has already 
been pointed out by Whitman (18iM), pp. 287 and 302) in the li^ech and salamander, 
and I suspect that the sensitiveness of the blind fish, as observed by Eigenmann 
and quoted by Whitman (1899, p. 303), may also be in the nature of a lateral-line 


Having reached the conclusion that the downward swimming of the fishes could 
be brought about by stimulating the lateral-line organs through slight mass-move- 
ments of the water, I next «ttempted to ascertain the relative iini>ortance of different 
parts of the lateral-line system in this reaction. I prepared one set of the fishes in 
which the lateral-line nerves were cut close to the pectoral girdles, thus rendering 
ineffective the hiteral-line organs of tlie trunk while those of the head were left 
intact. These individuals responded in all respects, so far as I could see, as normal 
fishes did, and I therefore concluded that the lateral line proper was not an essential 
part of this system of sense organs. 

In the second set of fishes I cut the fifth and seventh nerves of both sides, thus 
preventing the lateral-line organs of the head from acting. These animals always 
descended when the aquarium was shaken, but with noticeably less precision than in 
the cases of normal individuals. It therefore seemed probal)le to me that the por- 
tion of the lateral-line system on the head was more effective than that on the trunk, 
but as this experiment involved cutting the general cutaneous nerves of the head 
as well as the lateral-line nerves, the experiment is not wholly conclusive. 

Finally, in a third set of fishes, I cut tlie lateral-line nerves and the fifth and 
seventh nerves of the right sides only, leaving the left sides intact. These fishes, 
though a litth* sluggisli, reacted in an essentially normal way. From these three sets 
of exi)eriments I conclude that the lateral-line organs may be considerably reduced 
without seriously impairing the action of the system as a whole, though the portion 
of the system on the head is less easily dispensed with than that on the trunk. 

The .skin in Fundulu.s hefrroclitus. — While I was experimenting on fishes in 
which the lat:eral-line organs had been rendered inoperative I was at times puzzled 
by getting reactions that seeme<l contradictory to the general conclusion that such 
fishes were not stimulated by a slight movement of the whole mass of water. Occa- 
sionally on making the whole a([uarium move slightly a fish without lateral-line 
organs would swim rapidly to the bottom. On watcliing for instances of this kind 
I soon found that they occurred only when the fishes were close to the top of the 
water, and in fact were within the rang*^ of wav<^ action. When the whole aquarium 
was moved, even only slightly, th(» upper surface of th(» water wastlirown into small 
waves. These waves, as could be seen by tho motion of small suspended particles, 
extended only a few centimeters below th<* surface of the water, l)ut they estal)lished 
a r(»gion into which tlie fishes without lateral-line organs would not ascend, and 
from which, if overtaken by the waves there, they immediately escaped by swimming 
downward. As fishes without ears as well as without, lateral-line organs were stimu- 
lated by these surface waves, I concluded that in this instance the motion of the 
wat<M' must affect the general cutaneous nerves (touch). 

If the motion of surface waves is a stimulus for th(» general cnitaneous nerves, 
it would seem probable that currents in the water would also affect these nerves and 
that the ability of a fish to head up a stream might depend rather on the stimula- 
tion of its skin than, as Schulze lias implied, on tlie stimulation of its lateral-line 
organs. Fundulws is in a marked d(»gree rheotactic, i. e., it swims vigorously against 
a current, and I therefore resolved to test this fish to ascertain whether its rheotaxis 
depended on its lateral-line organs or not. Six specimens, in which the nerves to 
the lateral-line organs had l>een cut, were placed one after another at the oikjii end 
of a large glass tube through which a moderately strong <*urrent of sea water was 
flowing. All swam energetically up the tube, and, so far as this reaction was 


eoncerne<l, they were in no observable respect to be distinguished from normal fishes. 

U I Their rheotaxis certainly did not depend upon their lateral-line organs, l)ut was 

. undoubtedly the result of cutaneous stimulation. Unfortunatoly I was unable so 
to operate on other individuals that I could obtain active specim(»ns whosc^ cutaneous 
nerves were severed but whose lateral-line systems were int^ict, and hence the c)nly 
conclusion I can draw is that the general cutaneous nerves are stimulated by wave 

! ^ and current action an<l that this is sufficient to account for rheotaxis, but I can not 

state whether or not the lateral-line organs are also stimulated by tlnsse means. 

f 1 Concliisinrus concernhuj the laterdl-line organs and the skiti, — The observations 

? :'■ on Fundulus recoi*ded in the preceding pages give no sujiport to the view of P. and 

F. Sarasin that the lateral-line organs are to be regarded as accessory ears, for 
individuals in which the eighth nerves had l)<*en cut and in which the lateral-lino 
organs were intact did not respond to the sound-waves from a tuning-fork to which 
fishes with ears reacted with certainty. I have also set^n no reason to suppose that 
the lateral- line organs are especially connected witli the production of gas in the 
air-bladder, as suggested by Uichanl, or that they are particularly concerned with 
equilibration, as advocated by Lee. Since they are stimulated by slight disturbances 
in the wat«r that do not jiiTect the general cutaneous sense organs, I can not agree 
with Merkel in classing them as tactile organs. Their appropriate* stimulus is a 
slight mass-movement of the wat<»r, which nun- or nuiy not be vibratory, and which 
induces the fish to swim into deeper regions. This fonn of stimulus is of precisely 
the kind that was attributed to these organs by Schulze (1S7U), but I have not lK»en 
able to confirm Schulze's further opinion that curn^nt and surfa<»e wave movements 
stimulate these parts. Such stimuli certainly do affect the general (Hitaneous sense 
organs, but whether or not they influence the lateral-line organs 1 am unable to say. 


Although my studies were made almost exclusively on Fundulus heterorlitU'Sy I 
tested, as oi>portunities offered, other species of common fishes. Th<\se were placed 
without l>eing operated upon in th(» aquarium with the bass-viol string as a means 
of producing sound. Because of the mixed character of the stimulus priKluced by 
this apparatus and also because of tli(» fact that the fishes were not operated upon 
in any way, the results are significant in only one or two instances. 

Young mackerel, while swimming in the aquarium, always moved downward 
when the strinir w'as vibrated. The same was found true of adult nuickerel, but 
wliether this reaction was an ear or a lateral-line response was not detei'iniiuHl. 

Menhaden, after they became somewhat accustomed to the a([uarium, gave a 
sudden leap each tinu^ the string vibrated, but sliow(?d no tendency to descend. In 
this instance, too, no clew was obtained as to the organs stimulated. 

Three specimens of smooth dog-fish, each about 18 inches long, were test43d. 
When these fish were resting quietly on the stone bottom of the aciuarium, the 
vibration of the string would cause them to move their pectoral and pelvic fins, 
or ev(m begin swimming, but when they rested on some *} inches of cotton wool 
covered with a cloth to afford a deadened surface on the bottom of th(» aquarium, 
no reaction of any kind was ever obtained. Apparently the eai*s, lat<»ral-line 
organs, and skins of these fishes are not open to any of the stimuli produced by the 
vibrations of the bass-viol string and transmitted through the water, and tliey tlius 
differ markedly from the other fishes examined. 


These few notes serve to show that diflferent species respond very differently 
to the same forms of stimuli and emphasize the imjK)rtance of refraining from 
generalizations on the functions of the lateral-line organs and the eai-s in fishes 
before a considerable number of species have been fully examined. 


1. Normal FnnduJus heteroditus reacts to the sound waves from a tuning-fork 
of 128 vibrations per second by movements of tlie pectoral fins and by an increase 
in the respiratory rate. It probably also responds to sound waves l)y caudal-fin 
movements and by general locomotor movements. 

2. Individuals in which the eighth (auditory) nerv(»s have l)een cut do not respond 
to sound waves from the tuning-fork. 

3. The absence of responses to sound waves in individuals with severed eighth 
nerves is not due to the shock of the operation or to other secondary causes, but to 
the loss of the ear as a sense organ. 

4. FunduJus heterorlifus then* fore possesses the sense of hearing. 

5. The ears in this species are also organs of prime imi)Oi-tance in equilibration. 

6. Normal Finidulus heferoditua swims downward from the top of the water 
and remains near the bottom when the aquarium in which it is contained is given a 
slight noiseless motion. 

7. Individuals in which the nerves to the lateral-line organs have been cut will 
swim upward or remain at the top while the aquarium is being gently and noiselessly 

8. The lateral-line organs in this sjx^cies are probably stimulated by a slight 
mass movement of the water against them. They are not stimulated by sound waves 
such as stimulate the ears. 

9. Individuals in which the nerves to the lateral-line organs have been cut swim 
downward and thus escajx? from regions of surface wave action. They also orient 
perfectly in swimming against a current. Since surfa<*e waves and current action 
stimulate fishes in which the nerves to the lateral-line organs and to the ears have 
been cut, these motions must stimulate the general cutaneous nerves (touch). 

10. The vibrations from a bass-viol string when transmitted to water stimulate 
the ears and the lateral-line or<^nns of Fund id u>{. They also stimulate mackerel and 
menhaden, but not the smooth dog-fish, which responds only when in contact with 
solid portions of an aquarium subjected to vibrations. 

The work recorded on the preceding pages was done at the biological laboratory 
of the United States Fish Commission at Woods Hole, Mass., and I take this oppor- 
tunity of expressing my indebtedness to the Director, Dr. Hugh M. Smith, and 
to his assistants for much help rendered me. I am also under obligations to Prof. 
W. C. Sabine, of Harvard Univei-sity, for advice and assistance in connection with 
the sound-producing apparatus, an<l to Prof. F. S. Lee, of Columbia University, for 
friendly criticism and many suggestions. 



Bateson, W. 
1890. The sense-organs and perceptions of fishes: with remarks on the snpply of bait. Journal 
of the Marine Biological Association of the United Kingdom. New series, vol. 1 , pp. 
235-256, pi. XX. 
Bonnier, P. 
1806. Snr le sens lateral. Comptes rendos des seances et memoires de la societe de biologie. 
S^rie 10, Tome 3, pp. 917-919. 

1880. Le Specie del Gkinere Fierasfer nel GoUo di Napoli, e Regioni limitrofe. Fauna und Flora 
des Oolfes von Neapel. 2 Monographie, Leipzig. 76 pp., Tav. i-ix. 
1895. Ueberdie Function der unter der ELaut liegenden Canalsysteme bei den Selachiem. 
Archiv fftr die gesammte Physiologie. Bd. 59, pp. 454-478. Taf . vi. 
GOnthee, a. C. L. G. 

1880. An introduction to the study of fishes. Edinburgh, xvi - 720 pp. 
Hunter, J. 
1783. Account of the organ of hearing in fish. Philosophical transactions of the Royal S<K'iety 
of London. Vol. 72, pp. 379-383. 
Krridl, H. 

1895. Ueber die Perception der Schallwellen bei den Fischen. Arcihiv fiir die gt^samnite 

Physiologie. Bd. 61 , pp. 450-464. 

1896. Bin weiterer Versuch iiber das angebliche Huren eines Glockenzeichens durch dio 

Fische. Archiv ftir die gesammte Physiologie. Bd. 6;}, pp. 581-586. 
Lee, F. S. 

1898. The functions of the ear and the lateral line in fishes. American Journal of Physiology. 

Vol. 1, pp. 128-144. 
Leydig, F. 
1850. Ueber die Schleimkanale der Knochenfische. Archiv ftir Anatomie, Physiologie und wis- 

senschaftliche Medicin. Jahrgang ia50, pp. 170-181. Taf. iv, figa. \-H, 
1868. Ueber Organe eines sechsten Sinnes. Dresden. 108 pp., Taf. i-v. 
Merkel, F. 

1880. Ueber die Endigungen dersensiblen Nerven in der Haut der Wirl>elthiere. Ri>Ht<Krk. 214 
pp. Taf. i-xv. 
MOller, J. 
1848. The physiology of the senses, voice, and muscular motions, with the mental faculties, 
'franslated by W. Baly. London, xvu i- pp. 849 to 1419 *-32 f 22 pp. 
Owen, R. 

1866. On the anatomy of vertebrates. Vol. i. London. XLi 14-650 pp. 
Richard, J. 

1896. Sur les functions de la ligne laterale du Cyprin dore. Compt^^'s rendus des seances et 
memoires de la soc^iete de biologie. Serie 10, Tome 3. pp. 131-13:J. 
Romanes, G. J. 

1892. Animal intelligence. International Scientific Series, vol. 44. New York, xiv i 520 pp. 
Sage, D., C. H. Townsend, H. M. Smith, and W. C. Harrih. 

1902. Salmon and Trout. New York, x f417 pp. 
Sarabin, p., und F. Sarasin. 

1887-1890. Zur Entwicklungsgeschichte und Anatomie der ceylonesischen Blindwiihle Ichtliyt)- 
phis glutinosus, L. Ergebnisse naturfwissenschaftlicher Forschungen auf Ceylon. 
&1. 2. 2<J3 pp. Taf. i-ui. 
1870. Ueber die Sinnesorgane der Seitenlinie bei Fischen und Amphibien. An^hiv fur mikni- 
skopische Anatomie. Bd. 6, pp. 62-^, Taf. iv-vi. 
Whitman. C. O. 

1899. Animal Behavior. Biological Lectures from the Marme Biological Laboratory, Woods 

HoU, Mass., 1898, pp. 285-338. 



RIV1:R, iS()()-Kjoi. 


\iiinraiisi, UfiiUd Stales Fish Conniiiss'wn SttatHi r .llhatruss. 

V.r H. I'.Nrj— I ,;5 






Origin and method of the invcittigation 07 

General results of the in vostigatioriH Os 

The Sacramento River ait a Kalmon ntream 70 

The general life hi8t4^>ry of the Paeittc nalmoiiK 71 

Salmon eggH and milt T2 

Experiments on the vitality of Kpermatozoa 73 

Vitality in water 73 

Activity in normal salt solution 73 

Vitality in air 74 

"Watery" milt 74 

Amount of milt requirtKl in artificial fertiliza- 
tion 71 

Experiments with ova 74 

How to test fertilization 74 

Short exposure to water detrimental 74 

Effe<;t of exposure to air 75 

Fish slime not deleteriotis 75 

Fertilization in body fluid 75 

Immersion in normal salt solution 75 

To make artificial spawning complete 70 

An aid to spawn-taking 70 

Pry process of fertilization 77 

Killing the female before spawning 77 

Quality of bloody eggs 77 

Foamy eggs 77 

Granular eggs 77 

Eggs dead when spawned 77 

Eggs from dead fish 7H 

Spotted eggs 7M 

Yellow eggs 7k 

The critical period for eggs 78 

Fungus in the hatchery hO 

The alevin HI 

Under natural conditions si 

Observations on alcvins artificially reariKl Kl 

General account si 

Plan t ing alevins K3 

Enemies Ki 

Results of observations h5 

Thefry m 

Notes on younger fry WJ 

Planting fry from the haU-hery SO 

OlwHjrvations of a particular fry M7 

Enemies m7 

Migration of fry i« 

I n Olenui Creek Hh 

Battle Creek sUtion h8 

Balls Ferry station W 

Walnut Grove station W 

Ol»servations at Bcnlcia 91 

General iiivcHtigation during Is^jh tfr> 

Movements in estuary and liay IfJ 

EfTect of sea water on alevlnn and fry '/7 


The fry — < Continued. 

Change of color during migration 100 

Summary of oli»ier\'ati< »ns on migration 100 

Summer n*idcnts in the rivers 100 

Gcnenil ucfctunt 100 

r|»stn»ani movement 101 

Migrati(m during summer 102 

Summary on ntmiber and movements 105 

Growth in fresh water 105 

Gastric parasites 109 

Di>ea!<><l parrs 109 

Matiure male iiarrs 109 

Tenijierature notes 110 

Conclusi(ms from study of summer resi- 
dents 1 10 

FimkI of young »alnion 110 

Gcneml study of f«ssl in fresh water 110 

F<kk1 in brarkish water 117 

Classification of the inse<'t ftssi of young sal- 
mon 118 

The jieriod of growth 120 

The arlult salmon 121 

Migration 121 

I Do salmon return to their native streams 121 

The two runs c>f halmon 121 

Details of uiignition 122 

Downstream movement 124 

Relation iMrtween weather and migration 125 

Changes in salmon after enti-ring fresh water 125 

The allmentar)' (ranal - 125 

The skin 126 

Ijtuf* in weight 127 

C-omiiarison of the sexes 130 

iU^lative changi-s in fresh water 131) 

Two forms of males l:U) 

HenmiphnKlites ijfcj 

Relative numlsir of males and femalch 132 

Ri'lative weight 133 

Natural propagation 134 

Spawning habits 1^4 

Percent4ige of fertilization 135 

Mortality among ova 136 

Natural verwus artificial pro|>agatIon 137 

Injuries and dlseastw 13^ 

General effects of sini wning 138 

S|H-nt salmon 139 

DiscHisrs of the intestine 139 

Fungus i:« 

(till iiaraslt4's 140 

DiM;as<*<l ova ho 

i^rngth of life of fall salmon after reaching 

s|iA wnlng-grounds 140 

Ut-BtU 141 


Daily observations of tho migrations of tho yonng salmon in the main river 
were made at two stations, al)Out 325 mile« apart, from January to May, ISOl). The 
hatchery experiments were carried on during two seasons, 1807 and 1808, at Battle 
Creek hatc^hery. Twenty-five thousand eggs were hatclied at the Hopkins Seaside 
Laboratory at Pacific Grove during the winter of 1808-00, and the young were used 
in experiments testing tlie effect of sea water on alevins and fry. Kxperinn^nts in 
planting alevins and fry were made at Olema, Marin County, and at Sisson, Siskiyou 
County, in 1807 and 1808. Tlie investigaticms in 1000 and 1001 consisted of observa- 
tions on adult salmon at Pacific Grove and at various points on the Sacramento, 
especially at Black Diamond, Rio Vista, and Mill Creek and Battle Creek fisheries. 

Observations of the habits of the young were nuwle first by watching them in the 
water. This, of course, could be done only in the small streams, but was very suc- 
cessful in the work the first year at Olema. Specimens were secured by means of 
the ordinary Baird seine; the one most used in this investigation was 50 feet long 
and 7 feet deep; smaller seines, 20 and 15 feet long, were use<l in small streams; in 
some of the work in San Pablo Bay a net 150 feet long was used. Traps constructed 
to suit particular purposes were employed wlien? occasion reiiuii^ed. 

The work was carried on under the direction of the United States Fish C/om- 
mission, with the cooperation of the California Fish Commission, through Mr. N. B. 
Scofield, during the first two years. Mr. F. M. Chamberlain, of the United States 
Fish Commission steamer AlbatrosSy assisted in the work from May, 1808, to April, 
1800. Mr. A. B. Alexander, also of the steamer Albatross^ began the work at Olema 
in 1807. Much of the success of the investigation is due to the interest and counsel of 
Mr. J. P. Babcock, of the California Fish Commission; and Prof. Charles II. Gilbert, 
of Stanford University, has aided much in planning the work aiul in affording 
facilities ior studying the collections. 

The author is under obligations to the agents of the Southern Pacific Company 
at the shipping-points along the Sacramento River, to the various fish-dealers in 
Sacramento, and to the salmon-packing associations at Benicia and Black Diamond 
for statements of the catch of salmon at various places along the river and bays 
and for other courtesies; also to the officials of the United States Weather Bureau at 
San Francisco and Red Bluff and to the agent of the Southern Pacific Company 
at Sacramento Bridge for river statistics. The direct^)rs of the Hopkins Seaside 
Laboratory at Pacific Grove gave the free use of laboratory facilities for carrying 
on the experiments at that place. 


A great many points of more or less interest have l)een considered in this series 
of investigations, of which the following have the most practical value and deserve 
special mention: 

1. The original object of the investigation has been carried out in determining 
that young salmon should be released from the hatcheries soon after the yolk has 
been entirely absorbed and that they should not be released in the heailwaters late 
in the spring. 

2. A method has been found for removing and fertilizing the eggs left in the 
fish after artificial spawning, thus increasing tl.e take of eggs from a given number 
of fishes by about one-fifth. 

3. The site for a new hatchery, Mill Creek station, has been discovered. 


Other points of interest determined are: 

a. The spermatozoa of the milt are active for only 3 to 5 minutes aft^r the 
milt is mixed with water. 

h. Ova become incapable of fertilization after 5 minutes' immersion in water, 
and goo<l results can not be obtained after 1 minute. 

c. Ova can be exposed to air for half an hour provided they are kept moist by 
the ovarian fluid. 

d. Ova may be fertilized while immersed in the ovarian fluid, or in the slime 
from the skin, or in unclotted l)lood. 

e. Ova are not aff<*cted l)y immei*sion in normal salt solution for half an hour, 
and are capable of fertilization thereafter. 

/. Between the ages of and 10 days, when the water temperature is about 50", 
the embryo is especially' sensitive and liable to injury. During this period the eggs 
should l)e handled >vith the greatest care. 

g. Fungus is not a gi'eat ix^st at Battle Creek hatcherj-, probably owing to the 
considerable amount of silt carried in the water and deposited on the mat of filaments. 

h, Alevins have many enemies in the streams; fry but few. 

/. The frj- begin feeding and commence their downstream migration as soon as 
the volk is absorbed and thev are able to swim. 

j. The fry drift downstream tail first, traveling mostly at night and averaging 
about in miles a day. They ai*e 4 or 5 months old when they reach the ocean. 

k. A few of the later winter fry, about 10,0<X) to the mile in the Upper Sacra- 
mento, remain in the headwatei-s all summer, which is deleterious on account of 
slow growth. 

/. The food of young salmon at all places and seasons is insects, larval or adult. 

in, Salmon si)end from 2 to 4 years in the ocean. 

n. They usually i-eturn to the river through which they reached the ocean, 
becaus<» <luring their ocean life they do not get far away from its mouth. 

o. The later fall s^ilmon asc(»nd the Sacramento River at the rate of 4 or 5 miles 
a day, being alx)ut 05 days reaching Tehama from Rio Vista. The spring salmon, 
without doubt, travel faster. 

p. Salmon do not eat after leaving the ocean, and the stomach shrivels up to 
alx)ut a t4Mith of its normal size. 

r/. Salmon lose from 15 to 20 per cent of their weight in migrating, and from 10 
to 15 per cent more in spawning. 

r. The sexes can not be distinguished in salt water, but they differ greatly in 
fresh water. Tlu* males develop the long hooked jaw, the large canine teeth, the 
dei»p slab-sided body, and the color usually Injcomes more or leas I'Cildish. The 
females do not change in appearance except as is due to the loss of flesh, the 
development of the ova, and to the change in color from silvery to olive. 

s. The males vary more in size than the females and are of two forms, adult and 
grilse; the grilse resemble the females, but are much small(»r. 

/. The p<»rcentage of fertilization in natural propagation is high, probably about 
85 j)er cent. 

//. The injuries received in fi-esh water ai-e mostly due to exertions in spawning 
the last few ova. 

r. It is well known that all Pacific salmons die immediately after spawning 
once, and this investigation simply lH»ars out the fact. 



The Sacramento is a large river, navigable for boats as far as Red Bluff, which 
is 225 miles by rail from San Francisco. It is quite crooked, and the distance by 
water from Red Bluff to the Golden Gate is about 375 miles. The river rises in 
several small lakes in the mountains about 20 miles west of Sisson, in Siskiyou 
County, California, and for nearly half its length flows through a narrow canyon. 
The upper portion is a tj^pic^l mountain stream, with innumerable ix)ols and rapids 
and gravel beds, forming ideal spawning-places for the salmon, though it has not 
been visited by many of them during the past few years. Near the lower end of 
the canyon it receives Pit River from the east. 

Pit River is a much larger stream than the Sacramento above their union. Its 
lower portion lies in a canyon and except in size is similar to the I'^pjKjr Sacramento. 
About 75 miles above its mouth are the Pit River Falls, which, until a fishway was 
blasted out, were impassable for salmon. The upper portion of Pit River lies on a 
plateau and during the summer is a very unimport^int stream. 

The salmon that pass Pit River Falls spawn in Fall River, which enters Pit 
River a few miles above the falls. When seen in August, 1808, Fall River flowed 
several times as much water as Upper Pit River, though it is only 12 or 15 miles in 
a direct line from its mouth to its source. It is about 100 feet wide and 3 to 4 feet 
deep, flowing through a level plain and taking its rise in several large springs. 

Hat Creek, draining Mount Lassen on the north, empties into Pit River a few 
miles below the falls. It is a considerable stream, but its ascent is difficult for 
salmon on account of very steep rapids. 

McCloud River, draining Mount Shasta on the south, empti(»s into Pit River 
near its mouth. It is two or three times the size of the Sacramento River above the 
mouth of Pit River and is an important salmon stream, iiaird liat-i'hery is located 
near its mouth. 

The most important salmon stream of the basin, excepting the main river as 
noted below, is Battle Creek, which drains Mount Lass(»n on the west and enipti(»s 
into the Sacramento between Redding and Red Bluff. Battle Creek hat4*lu»ry is 
located at the mouth of this stream. 

A few miles below the mouth of Pit River, and just abov«» Redding, the Sacra- 
mento emerges from the canyon through which it runs from its source and widens 
into a broad, shallow stream, though the current continu(»s swift. Below Redding, 
for perhaps 100 or 150 miles as the river winds, it continues l)roa<l and shallow, with 
many short riffles and usually a gravel bank along one sidc^ In ordinary years 
when the river is in its normal low- water condition the princij)al spawning-beds of 
the fall salmon are in this portion of the main riv(*r, notably in th<» vicinity of Red 
Bluff and Tehama. In November, l(KK), the river was examined can^fully b<^tw(»en 
the mouth of Battle Creek and Tehama. P'ew salmon wen* s(»en until within a few 
miles of Red Bluff, but from that point on every riffle was coven^d with spawning- 
beds and dead salmon were everywhere abundant in their vicinity. Seventy-live 
dead fishes were counted at one time in the lower 100 yards of Mill Cre(»k and in the 
river within 50 yai*ds of its mouth. 

A few miles above Red Bluff the river cuts through a range of hills, and for 2 
or 3 miles consists of a series of rapids, the longest of which is known as Iron ( 'anyon. 
After passing Iron Canyon the river again assumes the character found at. and 



below Redding. Farther downstream the channel becomes deeper, gravel banks 
disappear, sand banks become less frequent, and rapids an^ wanting. Such is the 
character from Colusa to Sacramento. Below Sacramento it runs through a level 
country and for most of the distance is affected by tides. Th(»re are many sloughs, 
some connecting it with the San Joaquin. The Sacramento and San Joaquin rivers 
join as they empty into Suisun 15ay. 

The water of the upper part of Sacramento River and the upper tributaries is 
quite clear, and continues so until the mouth of Feather River is reached, from 
which point to the mouth it is very mudd3\ It is in the muddy water between the 
mouth of Feather River and Vallejo that the salmon for the markets are taken. 

The only species of salmon regularly frequenting the Sacramento River is the 
quinnat. The dog salmon is found occasionally, four specimens having l^een seen 
during this investigation. Mr. Chamberlain reports finding single specimens of the 
blueback and humpback at Baird in 1809. The humpback has also been reported 
by others. The only record of the silver salmon from the Sacramento River is that 
given by Jordan & Jouy (Proc. U. S. National Museum 1881). 


The salmon under consideration in this repoi-t is the Sacramento or quinnat 
salmon (Oncorhynchus tschaicyLschd), also known as Columbia River salmon, king 
salmon, and chinook salmon, and is the largest and most important of the five species 
of Pacific salmons. The others, in the order of their importance, are (1) blueback 
(O. net'ka), also called red salmon, rcMlfish, and sockeye; {"2) silver salmon (O. 
kifiuMi); (3) humpback salmon (O. (jorbiiscJia), and (4) dog salmon (O. keta). 

The Pacific salmons, as al)ove named, are different from tjie Atlantic salmon, 
which is related to the st,eelhead or salmon trout of the Pacific coast. One of the 
important charactei*s that separate the Pacific salmons from the steelhead and its 
relatives, the Atlantic salmon and the rainln^w and cut-throat trouts, is the larger 
number of rays in the anal fin — tho unpaired fin on the under side of the tail. The 
st4ielhea<l and it^s allies have fewer than 13 rays (usually 11) in this fin, while the 
Pacific salmons have more than 13, the number for the quinnat l)eing !(>.'' 

An equally fundamental, though physiological, difference lies in the fact that 
the Atlantic salmon and steelhead trout spawn several times while the Pacific 
salm<m, of whatever species, dies as soon as it spawns once. This is a very striking 
difference and its importance can hardly' be overestimated. A further difference 
lies in the habits of the young. The 3'oung of the Pacific salmon seek the ocean as 
soon as they are able to swim; their migration is accelerated by high water and 
retanled by low water, and they do not return to fresh water till mature. On 
the contrary, young trout do not s(?ek the ocean for several months after they are 
able to swim, low water is an incentive to migration, and they run back and forth 
between fresh and salt water seeking food. 

The quinnat salmon is found throughout the Pacific coast from Monterey J3ay 
northward, but is less abundant north of Puget Sound. At spawning time it fre- 
quents the larger streams, especially those with estuaries. The ])lueback salmon 
is the most abundant of the i^acific Siilmons, and is most numerous in Alaska. Its 
favorite spawning streams are those tributary to lakes. The silver salmon prefers 

a For other anatomical chara<-'ter«. se*» .Ti)r<lftii & Evorniann, FisbeM of North au<l Middle America, BuUetm 47, 
United States National Muneam. 



the larger streams, though not necessarily those emptying through estuaries. The 
other two species are of little importance. They spawn in any kind of stream, 
frequently in mere brooks that empty directly into the ocean. 

The Pacific salmon, of whatever species, passes most of its life in the ocean, and 
upon arriving at maturity ascends the rivers to spawn. Sometimes enormous 
numbers pass upstream together; stories of their being *' thick enough to walk 
across on" are oft^n told, but I have never seen them quite so numerous as that. 

The upstream migration occurs sometime during the warmer half of the year, 
the earlier fish going farther upstream. Spawning occurs late in the summer or 
during th(^ fall, and as soon iis it is completed the salmon dies. 

Most of the eggs deposited are eaten by other fishes, or are killed by being cov- 
eitid with sand and gi*avel. Those not destroyed hatch in from s(>ven to ten weeks, 
acconling to the temperature of the water. In the cold waters of Alaska they an? 
four or five months in hatching. It requires about six weeks more for the yolk-sac 
to be al)sorb(Hl, when the fry are able to swim and are ready for their seaward 
migration. Most of the alevins, however, are d(»voured by other fishes before they 
are able to swim. It is to prevent this great mortality among eggs And alevins that 
artificial propagation has been employed. 

The young salmon start downstream as soon as they are able to swim, and 
reach brackish water when three to fiv(? months old, according to the distance they 
have to go. Those from tlie vicinity of Battle Creek hatchery reach Benicia in 
about seven weeks. It is not known wiien they reach the ocean, but probably soon 
after. The variation in the time of spawning and hatching makes the p(»riod of 
migration very long. The fry from the summer run biggin passing Battle CrcH^k in 
September, and from that time until April following there is a continuous stream of 
young salmon, al)out H inches long, passing that point. 

Although a large majority go downstream ivs soon as they can swim, many, 
es])ecially those hatched in the spring, remain in th(^ pools in the headwat^M's all 
summer and fall. There were estimated to be from 7()() to 1,000 in each of several 
pools in the vicinity of Sims during the summer of 1S9S, and there were probably as 
many as 1(),0()() to the mile in that portion of the river. These remained in the 
headwaters until the first of December, when rains caused a rise in the river. 

'^i'lie food of young salmon in fresh water, at all times, places, and ages, consists 
of insects, either larval or adult. 


The following notes (»ml)ody obs(»rvations and experinu»nts madt^ at Battle Creek 
hatcher}' in 1S1)7 and 181)8. Although d(»aling largely with artificial propagation, 
they are not intended to give <^ven a gen(»ral account of the metho<ls of fish-culture 
as applied to the Pacific salmon. For such an account reference is made to the 
Manual of Fish-culture issue<l by the Unit^Kl States Fish Commission.'* 

For faciliti(»s placed at my disposal I am under obligations to Mr. (1. II. Lambson, 
superintendent of the stiition. Special acknowledgments are due to Mr. William 
Shebley, sup(»rint(»ndent of Sisson hatchery, and to Mr. Robert HadclifT, of Baird, 
who had charge of the spawn-talking operations in 181)7 and 181)8, respectively, and 
heartily coo]M^rate<l in the experimental work. 

RoiKirt United States Fish CoinmiHsioii 1K«7. im^^s 1 'MO. 

Bull. U. S F. C. IBM. lT.f«.p.,.H.) 

PLATE 11. 

•^V'^ " '<*=',■ . .:-Vl'^'. "■<■'"■.'-"■■ 


■ •■ '>*3^^ %^yj^^^:fm 

f ■: 








Vitality in irat^r. Experiment No. 1. — Milt was mixed with water until the 
mixture lH'(*aine of a milky api>eaRinee, which was the eomlitioii desire<l for fertiliz- 
ing the ova at the spawning platform. Small quantities of the mixture were taken 
from time to time and examineil under a compound microseoix*. After several 
repetitions of the experiment it was found that the si)ermatozoa remained active in 
the water from 3 to 5 minutes, the lenjrth of time varyinj^ slijrhtly in each exi)eri- 
ment. Eacli time the spermatozoa were phiced on the slide they were verj' active for 
alK)ut 3(^ Seconals, after which time nearly all became attache<l by the tail to the 
slide or cov(m\ The head ccmtinued to move for a few seconds hmger, but all 
motion ceased after 05 seconds from the time the milt was placed on the slide.}erimtnt Xo. 2, — Water was mixed with milt as in the al>ove (experiment and 
a small quantity used to fertilize eggs every half minute for >>^ minutes, with the 
following results, the* eggs lK*ing freshly spawned in each case: 

Timt* milt ba<I Ihnmi in the wiitiT. 


age of 

Time milt had l)H<»n in the water. 



U.2S> minntp 




4.5 minutes 


0.5 minute 

5 minutes 

5.5 minutes 


I minnte 

1.5 minutet» ^ 

2 minntcH 

ft.5 minutes 

2.5 minutOH 

7 minutet} 


3 minnU*H 

7.5 minutes 

3.5 minutes 

H minut4>s 

I m\x\xiXf» 

8.5 minutes 


Tlie exiK»riment was tried three timers, but the i*esults were practically the same. 
At one time a numlK^r of eggs were immersed in water taken from the top of the 
can in which the spawn was taken from the spawning platform to the hatchery. 
This water was white from the superfluous milt which had l>een spawned fi*om 
3 to 10 minutes. None of the eggs so tivated were fertilized. 

From these exiK»riments it will b«» st»en that the milt and eggs should Iw thor- 
oughly mixed while in the spawning pan and within W seconds from the time 
the milt is mixed with water. 

Activity in normal solt solution." — Milt was mixed with nornml salt solution 
until the liciuid was distinctly whitish, and a iM)rtion of it was at intervals poured 
over fivshly si)awn<»d eggs. After a short time the eggs were washed with fresh 
water. The jx^r cent fertilizer! in ea(*h case is given in the following tjible: 

Tim(> H< lint ion had lu'en sjiermatized. 

1 minute . 

2 minutes 
[ 4 minutes 

5 minutes 
♦5 minut*"* 
8 minuti's 

PercentAffe of fer- 










"Nonnitl salt siilutiou. (1.75 i>er eent n>mmou s.ilt in water. 


The results of this experiment are too varying to be of any practical valne. It 
seems probable that both si>ermatozoa and ova remain passive in the salt solution 
and that fertilization takes place only after the addition of fresh water in washing. 
Where a considerable quantity of fresh water was added and the ova mixed well with 
it befoi'e it was poured off the x)ercentage of fertilization was high. When the water 
was poured off immediat/cly and without mixing the ova well the fertilization was 
incomplete. Normal salt solution apparently jireserves the vitality of the sperma- 
tozoa longer than fresh water. 

VitnLity in air, — It was found that milt kept in an open, large-mouth bottle for 
24 hours fertilized 74 per cent of the eggs it was mixe<l with. Milt that had been 
so expose<l 48 hours did not fertilize any eggs, nor did that kept in a tightly corked 
vial for 24 hours. 

^^ Watery ^^ tnilL — Milt when taken from the fish varies greatly in consistency. 
That from some fishes is very thin and is known as wat(»ry milt. Experiment 
proves that it fertilizes eggs as well as any and that no larger quantity is needed. 

Amount of milt required in artificial fertilization, — In taking spawn it was 
the custom at Battle Creek to express the eggs from one female into a pan con- 
taining about a pint of water and add enough milt to make the water distinctly 
whitish. The amount of milt necessary for this varies, depending on the amount 
of abdominal fluid mixed with it, but is never less than 3 or 4 fluid ounces. This 
method gives good results and should be followed when there is an abundance of 
males, which is always true at Battle Creek after the first few days of the season. 
A smaller amount of milt, however, will suffice. Ninety-six per cent of the eggs from 
one female were fertilized by a tablespoon ful of milt; 85 per cent were fertilized by 
a teaspoonful; 35 and 57 per cent were fertilized by spawning fishes in the creek and 
letting the milt float over the eggs, which had been caught on a screen. 

Of course, all that is necessary is to bring a very minute quantity of milt in 
contact with each ^.gg, A single drop of milt if thoroughly disseminated through 
the water wouhl be sufficient to fertilize all the eggs from one female. In the experi- 
ments abov(^ noted the milt could not have l)een thoroughly mixed until after it had 
become inactive. It is not advisable to use less than a fifth of an ounce of milt to 
fertilize 1,CKK) eggs. More water is necessary where a small amount of milt is used 
in order to facilitate thorough mixing. 


Hov^ to te,st fertilization, — The quickest and surest way to det-ermine whether ova 
have been fertilized is to put them into a dilute (5 to 10 per cent) acetic acid. This 
can be made from commercial acetic by adding from two to five parts of water. A 
few after the ova have been placed in the acid the embryos turn whites 
while the yolk remains clear. The embryo can l)e distinguished in this manner 
within 15 hours after the ovum has been fertilized. In making this t^est for the first 
time it is best to make a comparative test with unfertilized ova that have been kept 
in water during the same period. 

Sliort e,rpof^ure to wafer detrimental. — A quantity of eggs were spawned into a 
pan of water and some were removed and spt»rmatized every half minute for several 
minutes, and for various periods up to several hours. The milt was, of course, 
taken fresh each time. 



Tlie rosiilts were as follows: 

Time f p^ had b«M>n in water. 

Percent- {i 
atre of 

Time egga had lK»eu in water. 

al of 







17 , 

4 minntefl 


4.5minntes. , .. 


5 minutes 

1.5 minutes 

5.5 minutes .. 

2 minu te» 


2.5 minutes 

6.5 minutes 

a mif^nt^-'s ... , ,, 

7 minutes ... 

3.5 uiinuteB 



The susceptibility of ova to fertilization decreases rapidly after they are placed 
in fresh water, and the milt can not be added too quickly. Fifteen seconds is as 
long as they should be in the water before the milt is added, and it is preferable to 
add the milt at the same time that the ova are spawned. 

On one occasion the eggs remaining in the Ixnly cavity after artificial spawning 
were removed by cutting the fish open, wliicli mixed them with much blood. They 
were waslied as quickly at possible and then spermatized. Only 11 per cent of the 
eggs were fertilized. In another experiment 85 per cent were fertilized in the blood 
without washing it off. 

Effect of exposure to nir. — A number of eggs were 8i)ermatized after having 
been exposed to the air, temperature 7C°, for various periods with the following 
results, the ordinary method giving fertilization of 99 per cent: 

Time ova were expose<l to air. 

12 minutes 
:*) minutes 
W minutes 

of fertil- 


Apparently there is no injury to eggs by an exposure to the air for half an hour. 
It must be noted that although the eggs were in an ojmmi pan they were practically 
immersed in the liquid from th(» body cavity. 

Fish slime not deleterious. — It is sometimes said that the slime on the fishes is 
fatal to the spermatozoa, an<l at some stations much care is taken to wipe the fish 
dry before spawning. To test the truth of the supposition, a pan of (*ggs were cov- 
ered with slime scraped from several fishes, and then spermatized without the use 
of water. Out of 174 eggs examined only two were unfertilized, which is as good as 
js obtained by th<* ordinary process. 

Fertilizntion in hodij jiuul. — At another time, 39*2 eggs were sj^ermatized "dry," 
and the milt entirelv wash(Ml off with normal salt solution before water was added. 
All but six, or 98.5 \wv e<»nt, were fertilized. This proves that water is not neces- 
sary to excite tli(^ activity of the spermat<^zoa, an<l that fertilization may l)e effected 
in the abdominal finid alone. 

Immersion in nor unit soft s( tint ion. — A quantity of egus were immersed in 
normal salt solution and at tin* end of certain periods were taken out and fertilized 
in the usual manner. Th(» results as shown in the table, while as good as could be 
desinnl, are scarcely In^tter than are obtained by the ordinary method, which gives 
a fertiliz^ition of 99 per cent. The value of the experimt^nt lies in the fact that it 
gives us a method of washing bloody eggs without preventing their fertilization, as 


will be noted below. By making a chemical analysis of the fluid of the body cavity, 
[I a liquid could probably be prepared that would be entirely passive and in which the 

ova could b<> kept for days. This, however, is unnecessary, as a saltness of three- 
fourths of 1 per cent gives a liquid sufliciently passive for washing out blooily eggs. 

Time tiggB were in salt Holution. 

of fertil- 

2 miniitcA 


4 miniiU^H 


6 minutes 


8 minut^H 


15 minutes 


25 minutes 


To make artificial spawning complete, — Even the best spawn-takers can not get 
all the eggs from the fish. Often the fish is not entirely ripe; but whatever the 
condition may be many of the eggs are (mtangled in the folds of the ovary and 
viscera and are not spawned. Under natural conditions the ovary shrivels up and 
does not obstruct the outward passage of the eggs. The number of eggs remaining 
in the fish after the artificial spawning varies from 200 to 1,5(K), depending upon the 
size and condition of the fish. I have found an average of 000 eggs reuiaining in 
55 fishes after artificial spawning. The spawning wjis done by experience<l men 
and could not well be improved upon. 

The average number of eggs taken from a fish in onlinary spawning is 5,000. 
This was the average during the season of 1807. l^y removing the remaining 000 
eggs tlie yield can be increased al>out 18 per cent. They can all b(^ removed only 
by slitting the abdomen from the p<^ctoral fins backward, but this allows a large 
quantity of blood to mix with them. It is possible to fertilize 85 per cent of these 
eggs in the blood, but in this case the unfertilized 15 percent have to bi» picked 
out of the hatching baskets, which would be a considerable expense if the plan 
were followed. But the blood can be removed from the eggs without any detriment 
to fertilization by wiishing them in normal salt solution ((me ounce of common siilt 
to one gallon of water). They can then be fertilized in the ordinary manner. 

. This method lias been used at Battle C.reek hat<*>hery since 1000 with satisfactory 
results, the loss with the "remnant" eggs being but little great(*r than with the 
ordinary take. By care in handling the loss need not be any greater, l^y such 
means the take of eggs (^an be increased from 10 to 20 pc^r cent without inc^reasing 
the cost appreciably.'* 

An aid to ,spawn-t<ikin(j. — It was found that fishes were much more easily and 
rapidly spawned after cutting the body walls across the* opc^ning of the oviduct. 
Unless the cut was made considerably in advance of the vent no perceptible amount 
of blood issued. A greater percentage of the ova were spawiuMl than if the gasli 
had not been made, and no eggs were broken in spawning, which is an imiK)rtaiit 
point. The shells or " shucks " from eggs broken in spawning are a great nuisance in 
the hatching basket, Ix^ing difficult to pick out and forming a basis for the growth 

«Prom his study of the physiolo|?y of the Sacramento salmon in IWCKJ, Prof. C. W. Greene, of the University 
of Missouri, has determined the amount of salt in the ovarian fluid to 1k» (».m jn^r <"ent, whi<"h. therefore, is the 
density of the solution that is normal for salmon ova, and should l»e use<l in washini; the 1>1(mk1 from ei^to*, rather 
than 0.75 i)er cent as used in the experiments here noted. A solutir>n of O.W per cent can l>e made by aildint^r U ounces 
of pure dry salt to 1 gallon of water. 


of fundus if llu'y aicMiot removed. It was found by eoiiiiting the numl)er of broken 
eg^s in several lots spawne<l in the ordinary manner that they averaged nearl}' 1 per 
cent of the entire take. 

In the ordinary meth<Kl of expiwssing the eggs they leave the oviduct under 
considerable pressure and strike the spawning pan with as much force as if they 
had fallen several feet. This manner of spawning, as alreaily seen, breaks nearly 1 
I)er cent of the eggs, and there may l)e many among those not broken that are 
injured; this nmy account for the* heavy loss of the fii*st day in the hatching-house. 

Dry (ff ferfilizatioiL — The method of fertilization used at ^iattle Creek 
in 1807 and ISOS was to spawn the eggs into a pan containing a little less than a pint 
of water, si)ermatizing them at the same time. They were then allowed to st^nd 
about 2^ minutes, when they were poured into a large bucket and gradually' wiished 
by adding fresh water. Basket Xo. of the table given in the notes l)elow on the 
critical i>eriod exi)eriment was treated in this manner. The eggs of basket No. 5 
were fertilized without any water, but otherwise were treated the same as No. 0. 
There was a difference of only 0.3 per cent in fertilization. There was a difference 
of 0.2 per cent between baskets No. 6 and No. 7, and they were from the same fishes 
and treated in the same way, so far as fertilization was concerned. 

The metho<l used at Battle Creek seems the Iwitter, as the eggs can l>e mixed with 
the milt more easily. A half minute, or just long enough to mix the eggs thoroughly, 
is an abundance of time for them to remain in the si)awning-pan. 

Killing ihe female before spdwning. — It has been clauned by some fish-culturists 
that killing the female before spawning causes deformed fry. Basket No. 1 (see 
table on p. 70) contained eggs from fishes killed by a blow on the head. There were 
not even so many deformities in it as in others. This method of procedure is not 
recommended, however, as green fishes would sometimes be killed, and their eggs 
therefore lost. 

Qualify of bloody eggs, — Occasionally a female has been injured before spawn- 
ing, and the eggs when pressed from the body wore mixed with blood. Eggs from 
three such fishes were kept separate; 7.7 per cent of the eggs died within five days; 
of the remainder, 2 per cent (.*3 out of 154) were unfertilized. The fertilization was 
about as good as the average, and a small amount of blood seems not to be detri- 
mental to fertilization. Several females were oi>ened aft«r spawning, and the 
eggs remaining were removed. The eggs were mixed with a great deal of blood, 
and only 85 per cent could be fertilized, so that a large amount of blood is detri- 
mental to fertilization, probably l)ecause clots of blood prevent thorough mixing 
with the milt rather than from any injurious effect upon the ova or spermatozoa. 

Foamy eggs. — Often the ovarian liquid becomes foamy as the eggs are spawned. 
It was not known whether such eggs were fertilizable. In the foamy eggs experi- 
mented with IM.) per cent were fertilized. 

Granular eggs. — The eggs from a certain small salmon, owing to the arrange- 
ment or sui^erabundance of oil globules, had a j)eculiar granular api>earance. Fer- 
tilization by ordinary process was IM) j)er cent; appai*ently healthy when 26 days old. 

Eggs (lead when spawned. — Occasionally eggs at the time of spawning have a 
dull, yellowish api>earance, and are evidently not healthy. They are always thrown 
away. Eggs of this kind from one fish were kept. Seventeen days after spawning 
30 j)er cent had died, and of the remainder 23 per cent were unfertilized. They 
were not kept for further observations. Such eggs were not found in 1898. 



Eggs from (lead fish, — On two occasions a ripe female was removed from the 
water, and after it had been dead 2 hours a few eggs were spawned and fertilized in 
the usual manner. A few eggs were spawned from time to time until the fish had 
been dead 34 hours. The following table gives the results: 

Number of hoiirn fltth had been dead. 

2 hoiiFH 

4 houFH 

6 hours 

8 hours 

10 hours 

1:J hours 

16 hours 

24 hours 

:M hours 

PercentaK*) <>' '«r- 




age of ogffs 
that dii»d 
within 10 






The first four lots of eggs of 2, 4, 6, and 8 hours wei-e kept 26 days and were 
apparently entirely healthy at the end of that time. 

At another time eggs were taken from two fishes that had died in the water. 
One had been deail 1 hour, the other over G hours. Of the first, over 97 per cent 
hatched and were healthy fry; of the other, 85 per cent. 

From the above it is evidently safe to take the eggs from fishes that have been 
dead less than 5 hours, and fairly good results can be obtained up to 8 horn's. 

Spotted eggs. — Sometimes a considerable number of eggs, a few weeks after 
fertilization, have a small, irregular white spot about the size of the head of a pin 
in the yolk near the surface. This does not mean that the egii; is about to die. 
Fifteen such eggs were put into a separate basket, and all hat<*hed as perfectly 
healthy fry excepting one, which died in breaking through the shell. The si)ot did 
not appear on the yolk-sac. 

Yellow eggs, — When eggs are nearly ready to hatch a yellowish fluid sometimes 
collects around the embryo. This does not affect them very seriously, as most of 
them hatch into healthy alevins. 

The critical period for eggs, — It is a well-known fact that at a certain stage in 
development, from about the sixth to the sixU^enth day, eggs are much more liable 
to injury than at other stages. When first taken they c^n be handled with com- 
parative roughness with impunity. At Battle Creek in 1807 the spawning platform 
was al>out half a mile from the hatchery. The eggs were hauh^d this distance over 
a road that lacked much of being smooth, yet the loss traceable to such handling 
was slight. Of nearly 700,000 eyed eggs sent from Battle Creek to Olema at one 
time, less than 300 were killed in shipping. They were hauled about 10 miles in a 
heavy wagon, were on the train some 15 hours, and out of the water 48 hours. At 
the time of shipment the eggs were 43 days old. But at an earlier date, when i\ to 
16 days old, such treatment would have killed every egg. 

For puri)oses of comparison 60,000 eggs from several fishes, fertilized in the 
ordinary manner, were mixed in a can at the spawning i)latform, and at the hatchery 
were equally divided between two baskets. The eggs of one of the baskets wi^re 
picked over daily regardless of results in order to remove the dead or addled eggs. 
The eggs in the other wei*e picked over in the same manner on the fii-st, third, 
twenty-second, twenty-fourth, and forty-fii'st days, and occasionally after that date. 
The former of these experiments was called No. 7, the latter No. 6. As a further test 



another basket, No. 1 , of 30,0(K) Qggn was picked daily. The loss of the baskets picked 
daily was from three to seven times greater than that of baskets not so treated. 

The following table indicates the comparative loss in baskets No. H and No. 7, 
and shows that eggs are very sensitive between the sixth and sixteenth days, and 
that they should l>e disturl)ed as little as jiossible during that i>eriod. Basket No. 
6 lost but 785 eggs from the fourth to the twenty-second day, while basket No. 7, 
which was picked daily, lost over S,()(X). After the twenty-second day the loss in 
this basket, No. 7, was only 013, while it was 1,3(59 in No. 0, the one not picked 
daily. The loss on the forty- third day was the result of being shijiped from Battle 
Creek to Olema, in which case the loss in No. G was nearly eight time« that of No. 7. 
It is evident, therefore, that daily picking t^kes out nearly all the weak eggs, but it 
is also strikingly evident that it takes out a very great many that are not weak. 

Table tthowing lomt of eggs in lHMke(.H No. and Ao. 7, taken y(PvemlH'r 15, 1S97. 

Age, dayH. 

Tern- Lo08. 


tore of -,- _ -_ .. li 
wat«r. No. «. No. • . 












































4:1 » 













Ajfe, days. 

1 pera- 




No. 6. 





0. 1. 















° F. 

! 47 



























16 i 




15 ' 


11 . 

4 : 

3 ! 

13 i 


9 1 

72 . 

24 i 


15 > 


14 i 










2,497 : 






<« Many killod while piddng ( went over baHkct but once). ''Not ho many killed while picking. 

c Died almost bl» as one i>urHon could pick them out. 

The table l)elow gives a summary' of the loss in four baskets; two. No. 1 and No. 
7, picked daily, and two. No. 5 and No. 0, pieke<l so as to avoid the critical iKjriod: 

Picked daily. 
No. 1. No. 7. 

No. 5 

No. 6. 

LoHBof eraH 

LoHH of alexins 

l>eformed fry 

Total 1<JHH 

Percentage of lo«8 

Percentage unfertilized 










9. 422 






2,497 I 
"112 ! 





a Tho lai^e 1o8b of alovina of No. 6 was caoued by accidental smothering. 


There wore »K),0(X) eggs in each basket at first. The eggs of basket No. 1 were 
from tish that were killed before spawning. Those of No. 5 were fertilized accord- 
ing to the dry process. Those of No. 6 and No. 7 were fertilized by the ordinary 
process nsed at Battle Creek. Nos. 1 and 7 were picked daily. Nos. 5 and (J were 
not disturbed during the critical period. The number unfertilized among the dead 
eggs was not determined l>efore the twenty -eighth day; therefon> the last item of 
the table is only relative. To get the true percentage of unfertilized eggs it would 
probably be about right to double that given. (It can be determined whether 
addled eggs are fertilized by putting them in about tS j>er cent acetic acid. The yolk 
of the e^g becomes clear, and the embryo, if there be one, turns white. A strong 
solution of common salt will clear addled eggs, but it also disintegrates very young 

Even after the critical period has passed, the most careful handling kills some 
fertilized eggs. Several tests show that from 10 to 20 jwr c<*nt of the loss after the 
critical period is in fertilized eggs that have been killed in handling. They should 
therefore be disturbed as little as possible. 

Fuiigus hi tlie luiMiery, — Fungus is a considerable pest in a hatchery, but the loss 
of eggs at Battle Creek traceable to this cause is very small. Numerous experiments 
were made in order to determine if the fungus would attack and desti*oy living eggs. 
Only on one occasion have I found a live egg attacked by fungus. This one had 
a few filaments of fungus growing on one side, and the egg had Ix^gun to dic^ where 
the fungus was attached. Whether the egg had start^Hi to die Iwfore the fungus 
attacked it or whether it was attacked first I can not sav, but all other observations 
and experiments indicate that the fungus attacks only the deml eggs. 

Fungus grows rapidly on dead eggs, and the filaments extend in all directions 
and entwine the adjacent eggs in a thick mat. This interrupts the circulation of 
the water and often smothers the eggs so matted. When eggs are smotheivd the 
embryo turns white before the yolk becomes addled, so that death from that cause 
can be distinguished. 

At Battle Creek, in 1897 and 1808, the baskets of eggs were gone over on the 
second and third days after spawning and all of the dead eggs picked out. They 
were not disturbed again until after the critical i)eriod, or about the twentieth 
day. This method was followed even where baskets (size, 23 by 15^ by r> inches) 
contained 40,000 eggs each. The number of eggs that died after the third day wjis 
small, and at the "breaking out," that is, the first picking after the critical period 
at about the twentieth day, the few dead eggs were found scattered here and there 
through the baskets. Each dead egg was covered with fungus, the tilaments of 
which had entangled the live eggs lying in its immediate neigh borhoinl, holding them 
together in a bunch. It was seldom that more than fifteen eggs were held togc^ther 
in such a bunch, and the dead eggs never exceeded three or four. 

Such treatment is not recommended for other stations, as the difference in the 
character of the water supply makes it necessary to carry on separate investigations 
for each station in order to determine methods of treatment. 

The reason Battle Creek is so free from fungus is that the water contains a 
considerable quantity of silt or dirt, and if the eggs aix? left undisturbed a couple 
of days they become covered with a fine sediment. This coll(H*t« on the fungus, 
which acts as a kind of filter, making of it a black nunldy mass and impeding its 


in^>wth. Clay or other dirt free from organic matter is often mixed in the water to 
destroy a growth of fungus on fry. 

While carrying on the experiments at Pacific Grove in January, 1S!K), when 
they were 3H days old the fry were attacked by a ver}' serious growth of slime, some- 
times called gill-fungus by fish-culturists. This slime was composed mostly of a 
microscopic unicellular animal with a silicious shell, belonging to the order Flagt-l- 
laUe, Some other microscopic animals and unicellular plants, such as diatoms, were 
present. The slime collccte<l on the gills of the fish and killed about two-thirds of 
them. They wen* treate<l with a '25 per cent mixture of sea wat«r, which was very 
effective. Those which had iK^en i*emoved to a mixture of sea wat^r l>efore the 
appearance of the slime were not affected at all. 

It must not be supposed from the statements given above that hatching troughs 
and baskets must never }ye touched during the critical period, nor that fungus is 
the onlv disease to which salmon ova and the alevins are liable. As has been 
stated above, I am not giving a general method of fish-culture, but an account of 
certain investigations. If the deposits of sediment on the eggs and troughs show 
traces of decaying organic matter, especially if there is a growth of slime on the 
Sides of the trough, everything must \ye cleaned immediately. There is no doubt 
tlmt microscopic plants and animals, such as bacteria and those mentioned above 
as having injured the fry at Pacific Urove, are very injurious to the eggs, alevins, 
and fry and must l>e scrupulously guanled against. 


The eggs that are not destroyeii in one way or another when deposited by the 
spawning tishes lie among the rocks, where they lo<lge and hatch in from 6 to 9 
weeks, the time dei)ending on the tem'perature of the water. The alevins also 
remain among the rocks at the lj<jttom for a few weeks, and their movements, slight 
though they are, expose them greatly to such fish(»s as the sculpin and trout. Dur- 
ing this time tin* yolk supplies them with what nourishment they need, but about 
four wecivs after hatching the quantity of yolk has become so small that it is not 
absorl)ed rapidly enough by the l)loo<l to meet the needs of growth. At this time 
also the alevin is able to swim a little, and it frequently leaves the l)ottom to snap 
at some floating object. Its movements are necessarily awkward on account of the 
unabsorln^d yolk, and it therefore attracts predaceous fishes. This is the most 
critical i)eriod in the life of the salmon aft<?r hatching. The yolk-sac disappears 
entirely at the age of 5 or <*> weeks, when the young are known as fry. ITiis is the 
age at wliich they begin feeding. 


Gencrdl (icconnt. — In December, 1S1)6, 855,000 eyed eggs were shipped from 
Battle Creek hatchery to Bear Valley hatchery in Marin County. Here they wen* 
hatched early in February, 1897, under the care of Mr. Frank Shebley, of the Cali- 
fornia Fish Commission. After the yolk-sac was al)SorlH*d, which was al>out 35 
days later, they were fed for a few days on curds of milk, and then, in the second 
week of March, were turned into Pai)er-mill Creek and its tributaries, Nicasio, 

F. C. B. IW2-6 


Olema, and Hatch(^ry eiiH^ks. Tho fry were strong and liealthy, and wen* turned 
into the stroains in the best of condition. The younjj salmon were watched day 
aftc»r day, and systematic observations were made of their movements, ha])its, 
enemies, and rate of growth. The work wiis first begnn by the United States Fish 
Commission and carried on until the middle of May. After a break here of thrive 
weeks the California Stato Comnussion carried it to completion. 

In the winter of 1807-98 eggs were again shipped from Battle Crc»ek to Hear 
Valley. This time the numl>er was increased to 2,()0(),00(^, necessitating the plant- 
ing of the alevins as soon as they l>egan to hatcli. All wen* planted l)efore the 
yolk-sac was absorbed. This, when taken in connection with the previous year's 
work, gave an opportunity to study the (?ouiparative efTect iveness of planting alevins 
and fry. The ah^vins w(»re transiKirted fi*om the hatchery t^) tlie i)lan ting-grounds 
in 20-gallon cans. It was possible to carry 4rO,(MK) alevins in such a can for two horn's 
at a temperature of 40" without loss, though 20,000 or 30,00i) to the can was the usual 
numlx'sr carried. Tliey were carried by wagon or rail as the case re(iuire<l. A 
wagon was found to be preferable, the jolting lK>ing an advantage, as the splashing 
in the cans kept the water w(^ll aerated. If the road was smooth or if carried by 
rail, the water hail to Ik> aerated frequently, and it was necessjiry to put fewer in the 
can. Although alevins appear to be very delicat<\ they stand transiH)rtation mueh 
l)etter than fry, and a much larger number can bo safely carried in each can. 

Paper-mill Creek and its tributaries, where the young salmon wei-e phintcKl, were 
never visited by the quinnat salmon. This was one reason that they were selected 
for the experiment, as any young of that siK»cies that w<^ might tlnd would be known 
to have Ixien planted there. The streams are rich iu aquatic insect life, alTonlingan 
abundance of food for the salmon fry. Trout and sculpins {Cotius) are the only 
i)redaceou8 fishes. The streams do not flow directly into th(» ocean but through 
several miles of brackish tidewat-t»r into Tonmles Bay, and the tninsition froui fresh 
to saltwat4»r is very gradual, removing the danger of the fry biding rushed too (quickly 
from river to ocean wat<»r. It was thought that if the fry could thrive* in these 
streams and jmss suc(»essfully into salt water it would be of a<lvantage to utilize 
coast hatcheries and plant in the smaller streams, where tin* young salmon would 
not be subjectcHl to their siii)posed enemies during the hmg journey from the Upper 
Sacramento to the sea. Tin* thing feannl in the experiment was that the streams 
would prove too short and that the young salmon would arrive* at salt water lH*fore t hey 
were ready t^) conform to the conditions of life they >vould hav<^ to encounter then*. 

Observations indicate that fry can 1k> as safely planted in PaiM*r-mill Creek and 
its tributaries as in the Sacramento River, and they reach the ocean six weeks (*arlier. 
If it is true*, as the exiwriments made at the Clackamas hatchery in Oregon indit'aU*, 
that most of the salmon return to fresh watc*r to spawn aftitr l)eing in the o(*ean two 
years, a diflTen^nce of one or two mouths in the time of i*eaching the ocean is worth 
considering. If the full growth is attained in 24 to 'M\ months, the average gain in 
weight is from 12 to 16 ounces a month. As the gain is necessarily slight at fli'st, it 
must be much more than a jiound a month later. Any extension of time for living 
in salt water is an increase of the rapid-growing perio<l, as the early p(»riod of slight 
increase in weight must l)e passed through in any case. This argument holds good 
only on the supposition that the individual would leave the oceau in a i)arti<*ular 
month. But the groat variation in the time in which the Sacramento salmon leaves 


tne ocean makes it quite certain tliat the lime is determined by other influences than 
the season. If so, it is doubtful whether those planted in Pai>er-mill Creek would 
have any advanta-ge over those i>lanted in the Sacramento River. 

PlaniiiKj aJvvins. — Alevins on l^eing lil)erated in swift wat^r swim frantically 
and scatter in all directions as they are swept downsteam. Most of them seek the 
bottom and crowd into crevices between the pebbles or get into quiet places under 
or iKdiind large l>owlders. Others find their way into still water along the edge of 
the stream, where they remain exposed io view. In moderately swift wat<»r some 
find a Icxlging- place on the bottom or near the shore Ix^fore the}' have been carried 
a hundred feet downstream, and it has to \)i^ very swift water that will c^irry them 
a hundred yards. For several hours aft^r lx*ing plant-ed in swift water many of 
them keep moving alK)ut. Often the place where they first lodge is too much 
exposed to tlie curivnt, and they are rei>eat'e<lly swept downstream, lodging here and 
there for a few moments, until they finally reach a quiet place where they can stay. 
After a few hours this moving al>out ceases and they remain quiet, retaining 
their places for at least several days. In one instance G0,()0() alevins were lil>erated 
on a very swift riffle in Sacramento River 2(X) yards above a quiet pool. The riffle 
was shallow, at no pla(*e over a foot deep, but so swift as to make it almost impos- 
sible for a pei"son to stand. The alevins all found shelter l)efore they were carried 
a third of the distance to the pool. On visiting the riffle a day later none could be 
found much over a huntlred 3'anls l>elow the place of planting and none was found 
in the pool below, which was seined thoroughly. All had found sheltered places 
and had ceased to move down with the current. When alevins are plant-ed only a 
few yards alK)ve a pool, even in moderately quiet water, large numbers will drift 
into it, where they remain if the}' are not eaten by trout or other fishes. 

When alevins are liberated in a pool or pond they at first scatter out near the 
surface, but soon settle to the bottom, where they keep up a constant wriggling of 
the tail and pectoral fins. Within a day or so they collect in bunches, appearing 
as brilliant salmon-colored blotches over the bottom of the pond. The constant 
motion of the individuals stirs up the silt until it is washed away from them and 
each bunch i*ests on the solid Ijottom. Alevins in a hatchery, by this constant 
motion, keep the hatching-trough free from sediment. 

During a fri»shet at the California State hatchery at Eel River a thick sediment 
of sand was washed into the hatching-troughs and came so fast that the alevins, 
just hatched, were unable to keep it from settling. It covered the bottom of the 
troughs to a depth of '2 inches, l)ecoming hard and compact. The alevins, instead 
of being covered, were found above the cement-like deposit, and none of them had 
been lost. This interesting incident demonstrates their ability to keep from being 
covered by sediment during a freshet, whether they be in jxHils, ponds, or troughs. 

Alevins in the pond at the Bear Valley hatchery Ix^gan swimming about in 
schools before the yolk-sac was entirely absorbed. The presence of predaceous 
fishes might have caused them to do otherwise. 

Enemies. — When alevins are planted on riffles they are inclined to congregate 
in eddies and sheltere<l places behind bowlders. In these places several thousand 
of them may Im? found huddled together in a bunch plainly exposed to view. They 
are not very shy at this age and do not appear to tr}- so much to get out of sight as 
to get out of swift water. The brilliant salmon color of the yolk makes them very con- 




spieuouH, espeHMally when they eolleet in bunehes. One »ueh ])un(*h in Olenia Creek, 
in which there w^ere four or five thousand, could l)e plainly seen from a point 50 yards 
distant.. In such instanees they are very much exposed to the ravages of du(*ks 
and j^eese — both tame and wild — herons, cranes, and other wading birds, and ev<»n 
of hogs, to say nothing of the fish in the stream. Alevins are very tempting mor- 
sels, and there is scarcely an animal that will not eat th<Mn when given a chan(*e. 
However, most of th<? ah-^vins (?an get. into crevices l^^tween and under the i)ebbles 
and bowlders, where they are much safer from the attacks of the fish that may Ik* 
in the stream than they would Ik» if they were in the pools or quiet places. 

Observations were also made on planting alevins in Sullaway Creek near Sisson. 
This is a very favorable stream in which to release young salmon, so far as i^reda- 
ceous fishes are concerned. The only fishes of the stream are rainl)ow trout, scul- 
pins, and quinnat. salmon parrs remaining from the season before. The plants were 
made on the riffles, but as t,h(»se were all rather short, many of the alexins drift^nl 
into the i>ools. They were lil)erated in the morning, and in the afternoon the pools 
were seined with a small-mashed net. The lish caught were examined to find to 
what exti^nt they had eat<»n the alevins. The ivsult is shown by the following table, 
which gives the length of each fish examined and the contents of its st<mia(*h: 

» I 



SpooioH ex- 





Other nuiteriul in 
stomach. ' 

1 Hnuill iiobblo. 
« in«H*t iar\'K». 

1 ' 

SiMH*i(>h e.K- 

Size, in 





Other material in 


. SculpiuH 





1 caddiA larx'a. 







:: inMH't larvir. 


IiuM.M*t larvu?. 







Salmon itarrs. 





1 iniMX't larva. > 






















1 iuHwt lar\'a: 1 
water bu^. 













2 winK^Nl inwM'ts. 



In the case of the largest 1 rout, (J inches lonir, 7 of the 17 alevins were in its thi*oat 
and mouth. It ha<l evidently gorg<»d its(^lf to the limit. 

In all cases wh(»re salmon pan's had eaten two alevins, the tail of the* second 
remaiiMMl sticking out of the mouth, their stomachs IxMiig large enough toacrcmimo- 
date only one. Th(» sculpins also had gorged tlu'iuselves in tlie same manner. All 
of tlH» tish caught were examined an<l only thive had not eaten alevins, InMug t<H) 
small. Tliree alevins had l)<»en <lisgorge<l by some of tlie tish. Kvidently ah'vins 
are a favorite food for trout, sculpins, and salmon parrs; and wh<»n they remain 
exposed to su(*h enemies from 2 to \ w<»eks, it is a wond<M' that any eseap<'. 

Alevins planted in the Marin County .stivams in ISOS mot evon a woi-se fate. Here 
the trout ar4» more numerous and largtM*. Tin* sculpins are also larger and more 
abundant. TIum'c w<»re no salmon parnsto f(H»d on al(»vins, but thei-4» were myriads ()f 
sticklebacks, whicli, tliough unable to swallow an alevin, kilU'd many by nibbling at 
the yolk. The only other tish in th(»se streams was tin* roa<*h (Jintilus stjinmrfrirn.s), 
which as far Jis could l>e learned did not fe4»d on \\u^ aU»vins. Four is a moderate 


estimate of the average nuiiil>er of alevins that a trout will eat in a day, at which 
rate each trout would destroy about 150 before the absorption of the yolk-sac; and 
1,0(X) trout would destroy 150,000 alevins. The lesson is obvious. 

Just here it may 1x3 well to state that in 1807, although only 150,000 try — not 
alevins — were planted in Olenia Creek, large numbers of them were yet to be found 
in June following, and ([uite a number in August. In 1898, 850*000 alevins were 
planted in Olema Creek, and in June following there was a smaller number left in 
the stream than was found in August the year liefore. There are two ways to 
account for this. One is that the alevins were washed out to sea before they began 
swimming; but it is more probable that they were eaten by trout and seulpins. 

In the spring of 1898, 7,0()(),(K)0 salmon were planted at the hatchery on Battle 
Creek about two weeks before the yolk-sac was absorbed. Although trout are not 
numerous there, the stream swarms with seulpins {Cotfus gidosns)y salmon fry 
remaining from the season before, Sacramento pike {Ptychocheiltis gran(lis)y black 
pike ( Or/7io Jo/j microlepidofus)^ hitch {Larinia exUicanda), split^tail {Pogonichthys 
7iiacrolepulotus)y and suckers {Catosfomun ocvidentalis). All of these, though they 
do not feed exclusively on animal matter, take salmon eggs and alevins when they 
can get them. The Sacramento pike is very destructive to young fish. The split-tail 
is the most numerous sjKicies, and lives on salmon eggs during the spawning season. 

Each day while the hatchery was in operation the bad or addled eggs picked 
from the hatching baskets were thrown into the stream. Usually the^'^ were thrown 
into a small brook near its entrance to the creek. In a very short time after 
emptying a can of eggs the split-tails always began to appear, running in from the 
creek. In a few minutes the water would be alive with them, almost a solid mass 
tumbling one over the other, splashing the water and crowding each other in their 
frantic efforts to get the eggs, until some were foi-ced into the mud at the edge, 
while othei*s were lifteil upward till their backs or bellies were out of water, or one 
might get into a vertical position >\ith its head or tail out of water. Frequently 
one would gorge itself till throat and mouth were so full that the passage of the 
water over the gills was shut off and it suffocated. It usually i-equired about 5 
minutes to consume 5 gallons of eggs. 

Alevins ai-e almost as helpless as eggs and fully as pahitable, and there can be 
little doubt of their fate when planted in such an environment. 

Re^sults of observations. — The egg and alevin stages are the periods in the life 
of the salmon when the cai*e of the iish-culturist Ls most needeil. The art of taking 
and caring for the spawn has been so perfecte<l that the loss in hatching need not 
be over 10 per cent, and is often less. The loss of alevins, if they an» retaine<i in 
the hatching-troughs or nursery i)onds, need not b<» over 2 per cent. If the young 
are planted during the alevin stage, the loss is very great. If large numbers of 
alevins are released in unsuitable places, where the lK)ttom is comparatively free 
from stones, and where such predaceous fishes as the split-tnil and tmut alK)und, the 
loss may even be greater than if the pan*nt salmon had been allowed to take their 
natural course in spawning. 

Young salmon should never Ik* planted until th(» yolk-sac has entirely disap- 
peared and their swimming power has fully developed, (»ven though the}' have to be 
fe<l a few days. There is no advantage in holding them after this time. 



PItinting fry from the hatchery. — Fry are tranaportfd from llic hatchery to the 
sti-eamu in thcHamo manner an already desorihed for alevint), but it is not practicable 
to carrj- over 10,000 in a can, even for a short diatauce. They require more eare 
than aleviim, it being necessary to aerate tlie wafer constantly. 

When fi-y are liberated in running wat«r, Ihey imm«<1ialely head iiiistreaiii and 
try to stem the cun-«nt. ()win« lo tlieii' bcinjr more or Ivsn faint fnnn conlinenivnt 
in till! can the current nearly always <iarries them downstn-am a short disl^tnt-e, but 
they soon find their way into the more qiiiel water along the edn<' of the stn>ani, in 
the eddies or quiet pools, or among the stones at the iMittoni. S<ime even move a 
few yards al>ove the place of planting l>t'fore they come Ut ivst. On gaining quieter 
water they rest themselves, moving only enough lo keep from driftint: downstream. 
When ID such imsitioii they l>egin feeding <tn anj' partii-les of ftxid that' float within 
their vision, often snapping vicionsly at insects half as large as themselves. 

In a small stream tlien- is no marked tendency of tin- fi-y to form schools, eiK-h 
appearing to act indeiKsndently; but in a larger sln-am, ami esiK'eially in the largit 
pools, they often swim about in schools. II apjH'ars, too, from our observations 
in the t)acramenlo, that they run in schools after gaining the main river in their 
migration to the sea. 

After planting, the frj- soon Itegiu to drift downstream from one resting-place 
to another. This movement in small streams is not in schools. If many are plante^l 
at one place the movement downstream is quite rapid, and within 2i hours they will 
be scattereil evenly ahmg the stream for over a mile Ik'Iow the jdacc of planting. 
The ntovement, though marke^l in the daytime, is more general at. night. In one 
instance a screen was placed across a small stream a ((uailer of a niih' below where 
50,000 fry were releaseil. Although but few reache<l the scri.'en that day, the 
following morning apparently every one had reached it. (Mher observations have 
shown the same thing. Muddy water hastens the nuivemeut downstream, as does 
also high wat.ur, which is usually muddy. 

In Hatchery Creek, in Marin County, 150, (HK> fry 10 weeks ohl were released. 
They gradually s«'atlered dowuslream, floating t-ail fli-st. In four or live days they 
were about evenly distributed along the creek for 1^ miles ImjIow the hat^-hery. At 
tlie end of this time a net with a 10-inch cireularmoulh was plat-ed in the current 
in the daytime with month npstreiuii. In one lionr :(n (n- 40 fry were eauglil. This 
illustrates well the decideil movement downsti-eam after planliiig. 

When released in a large pool or pond the fiy collect in schools immediately 
and travel toward the inlet. 

In 18»8, I50,(MK» alevins were placed in a jwnd at the Hear Valley hatchery. 
These remaineti in the pond without l>eing fed nntil four wc-eks afler the absorption 
of the yolk-sac, vVs it hail but CtHi square feet of surfaci- an<l was only :.' or ;{ feci 
deep, there were obviously too many in the pond to do well without l>eiiig fed. As 
would lie expect^'d, they grew but little, though few, if any, ilied. Al llie end of 
four weeks all were very nearly of one size — 1.4 inches long. Those of the sjime 
age in the creek a mile l>elow the pond varied from 1.5 to 1.1> inches; s|)eciTiicns 
from (Jlema CIreek only two wtHsks ohler weiv from 2 to J.4 inches long. 


At any time dilrinj^: the four weeks that the fry were so crowded in the pond they 
could have gone out, as the overflow trough was unobstructed. Very few if any of 
them did so, however. Indeed, it wa« difficult to get them to go out at all, very few 
esciiping till nearly all the water was dra\*ni off. As soon as they came near enough 
to the overflow to feel the course of the current they would dart back into the pool 
again. It has often been noticed that fr}- have an aversion to going over a water- 
fall or swift rapid. The observations at Sims during the summer of 1898 indicate 
the same thing. On account of this, fr}' should not be planted above falls or swift 
rapids, especially in small streams, as it is desirable that they should move down- 
stream as soon as possi])le. 

Observations of a particular fry, — Fry were observed daily from September 18 to 
Octol)er .*^, 1000, in a pool between a rock and the shore in Battle Creek. The pool 
was about 18 inches across, 4 feet long, and 2 or 3 feet deep. There was but one 
fry until the 25th, when another appearetl. It is probable that only two individuals 
were seen during the observations, though we can not l)e sure that such was the case. 

When first seen the fry swam near the surface, but after a few days it remaine<l 
a few inches below. It stayed most of the time in the rather strong current, and 
was continually snapping at minute floating objects. When swimming near the 
surface it made from two to ten strikes a minute. Observations could not be made 
so easily after it began swimming deeper. It was seen to make at least 150 strikes, 
but each time whatever was caught was immediately ejected. Apparently it had to 
make a great many efforts l>efore finding anything edible. 

One of the fry was seen to leave the pool and resume its migration. It had 
been in the lower portion of the i)ool all day, and as evening approached allowed 
itself to be carried down into the shallow and swift water of the outlet, always 
keeping its head upstream. Several times it was carried halfway through the out- 
let, but darted back into the pool. Once it got entirely through the outlet and into 
the dei^p water below the rock and then darted bivck, but finally it was carried out 
into the main cuiTcnt, tail first, and was lost sight of. 

Enemies. — Asalrea<ly stated, 855,CHK) young salmon were planted in the streams 
of Marin County, Cal., in 181)7, after Imving been kept in the hatchery until the 
yolk-sac was absorbed aud they had iK^gun to f(»ed. In order to determine to what 
extent they were preyed upon by the other fishes of the stream, large numbers of 
trout and a few sculpins were caught and examined, lx?ing the only fishes in the 
stream that could Ije suspected of eating salmon fry. Beginning at the time the 
plants were made and continuing for three wec*ks, 30 or 4() trout, ranging from to 
10 inches in length, were daily caught and examined. In not one instance had a 
salmon been eaten. The only fish eaten by them was the small minnow {Rutihis 
synimetricus), and no more than 10 of these were found in al>out 70() trout examined. 
Of the sculpins {Cottu.s yulosu.s)^ only 25 of size large enough to eat a salmon Hy were 
cauglit. None of these had eaten fish of any kind. 

In 1808, after the 3'oung s^ilmou planted that year had al)sorl)eil the yolk-sac, a 
number of trout were examined. None were found to have eaten salmon fry. 

On one occasion a small i)ool 8 feet across an<l about IS inches deep was seined. 
Over KK) young salmon were caught, averaging 2.1 inches in length. Along with 
them alx)ut a dozen trout from 6 to 8 inches long wen* tak<Mi. It would seem that 
if ever trout ate young salmon it would be here. These trout were examined, and 



it was found they had eat<*n only caddis larvae and periwinkl<^s. On the Upper 
Sacramento River I have examined many trout taken whih* the stream was full of 
the small salmon fry, but have never found that they had eaten younjr salmon. 
The same is the ease witli the sculpin, and these are the only fishes to be feared in 
the Upper Sacramento. Farther down stream many of the smalh»r Sa<*ramento pike 
have been examined, but none of them were guilty of eating young salmon. 

In the spring of 1S9*J, while observing the migrati<m of 1hesalm(»n fry on the 
lower Sacrament^) River by means of a fyke-net trap, we occasionally caught (*at-fish 
along with the 3'oiing salmon. In every cas(? it was found that tlu» cat-lish had 
eaten salmon fry. Their capacity for young salmon was great t»r even than that of 
the trout for alevins. Several (*at-fish i) inch(»s in length were found with over 0() 
salmon fry in their stomachs, and one of this sjime size* had eaten Stl of the fry 
which averaged a little over l\ inches long. To d<^t<M'mine whethiM* the cat-fish 
captured the fry (mly while in the bag of the net we caught nearly /in with hook and 
line. The stomach of none of them containe<l a young salmon. Thus it is evitlent 
that the cat-fish likes salmon fry and wouhl catch th<Mn regularly if it could. It is 
too sluggish a fish, however, to catch salmon fry under (»nlinary circumstances. 

The only other fish at all likely to pr(\v ui)on the young salmon in fresh wat4»r 
is the striped bass {Rocrus Uneafu.s), which is found in the lower river and in large 
numters in the brackish wator of Suisun liay. It is also found in San I^iblo and 
in San Francisco bays. I have no information on the subject, except that tlie 
striped bass preys to a large extent on the carp in the sloughs of the lower rivers 
and in the salt or brackish water feeds almost exclusively on small (M'abs. It is 
significant, however, that both striped bjiss and salmon aiv increasing in numbers 
in California waters, the former enormously, and it can not, therefon*, be very 
detrimental to salmon. Young pike, suckei's, and split -tails aiv abundant in the 
waters inhabited by the bass, and all are sluggish in comparison with the Siilnioii. 
It would seem that young salmon would l)e the last fish upon which tln\v wouhl prey. 
A young salmon is very active and strong and much moiv shy than ev(Mi a trout of 
same size; after it luis l)egun to swim about and feed it is perfectly nh\o to tak<» care 
of itself, and the number kille<i by enemi(»s in» Sacramento is very small. 





In Olenid Creek. — The fii'st year at Olema 15(),(HM) fry, and the sectmd y<»ar 
S5(),(H)() alevins, were nde*isc»d in Olema C-reek. The stream was s<»in<'d about a 
month after the fry wen* i)lant4*d in 1SJJ7, and in ISDS about a monlh after th<^ time 
when the fry should have lH»gun swimming. Very few young salmon were takcMi in 
either year, and the results show that over 'Jo iM»r cent had left th<»stn»ani within 
the month. 

Batile Creek station. — The observations in Battle Creek were mado while* we 
W4M-e engaged wit h the hatchery experiments during October and Nov4MiiImm\ isiis. 
In obtaining datia concerning the 3'oung salmon we used a oo-foot seine such as was 
(employed in nearly all of the investigations; but. the most im|M)rtaiit devicM* {'or this 
work was a trap which caught the young salmon as they were going downstream. 
The trap was made by sewing a piece of lin(^-mesh(Hl webbing a<*n)ss the mouth of 
the bag of a .*K)-foot seine and fixing a funnel to extend baek into the bag fn»ni the 
middle of the webbing. It was set in a strong current just Im»1ow the upper ra(*k at 



the Battle Creek fishery, with the wings extending obliquely upstream, their ends 
being about 10 feel apart. The fish were deflected by the wings to the middle por- 
tion of the net, and found their way through the funnel into the bag. No effective 
means could be devised to pi-event the funnel from l)ecoming choke<l with leaves or 
other trash, which often happened within an hour or two after the net was set. 
There were many adult salmon Ix^low the rack, and the}' often tore the net with their 
teeth and frequently got fast in the funnel. Part of tlie time the net was set during 
the day, more often <luring the night. Sometimes it was si»t for only an hour or two 
during the night. The following is a record of the catch, showing the date, the time 
of day, and the numlKjr and size of the fry taken: 

Record of salmon fn/ taken in trap at Battle Creek ftsJu^, Oct. 7 to Nov. So, 1S98. 







.wo. 1 





1.6 1.7 


4.0 4.3 




4.8 .5. + 1 6.2 



5 ... 
1 .... 

3 1... 
17 ; 3 

! o'.... 













5 .... 




! :::: 



.... 2 


. 10 2 












H 2 
' .... 




i 1 

« .... 

11 .... 

(1 .... 





.... 4 .... 
.... 1 .... 





5- 9 1), m 





*» .... 

4 ... 

B ..... 
tt .... 
(» .... 




1 .... 

4 i .. 

1 .... 


Nov. 12 

4 .... 


2tl 2 










H-9 p. m 


4 >j a. m 

12 la. in 


1 2a. m 



17 .. 
H3 12 
« . ... 

.![ . 





27 1 







2 Wu.m ... 


1-2 a. m 

H liji.m 




4«» .. 
24 .. 

. 2:^ 



5:i 1 







NoTK. Th«* inimlKTM in the (M»iumu ht'aded 1.5 ♦ and 5. i- indicate th(>nDml>ertak«^n that wen* aliont 
l.."> inrht's long, or aUmt 5 inrlu«s long, a.** th«.»ra«.» may l>e, but wen* releese<l without meaj«uring. 

From the preceding reconl it will l>e seen that all of the fry (not including the 
pan's) were practically of the same size, 1.5 inches long. Of the .'522 fry examined, 
only two were over 1.0 inches long, one being 1.7, the other l.s. The 1.5-inch 
specimens had just absorlxnl the yolk-sac. Indee<l, there was often a small amount 
of yolk remaining in the body, although the sac had disapi)eared. The size of these 
specimens shows that they l)egin their downstream migration as soon as they begin 
swimming, or at the age of six weeks; their continuing the siime size 4luring the 
two months shows that practically all start downstn»am at the siune age. If part 
of them had h«»ld l)a<*k for two or three weeks, this would have l>een indicated bv 

7 ft 

a greater variation in size. 



The record also shows that ordinarily the yotin)^ Baliuon travel at night. The 
trap was so set that they could not have avoirtetl it had they traveled during the 
day. That they can ho caught during the day is proved by their being taken in the 
ojien tow net set in Hatchery Creek, as noted al>ove, under "Planting fry from 
the hatchery." Fry were soon quite often lying in aptwl near tlieshore during tlioday, 
and were seen to rise to small insects that liglitWl on the water. Tliej- probably feed 
more during the day, which makes their migration slowt-r, or stops it altogether. 

On Novomljer 30 there wa.s a rise in the ereek and the water was muddy. The 
catch from S to !> a. m. was larger than the average niglil e^it'Ch at other times, showing 
that high and muddy water induces salmon fry to travel during the day. This fact is 
borne out also by the work at Walnut ttrove, an account of which is given )»olow. 

The gi'eat variation of the cjitch when the net was set for an h<nir or two during 
the night indicates that they travel in scltools. 

A trap similar to that used in IrtaSwas set in Battle Creek in liKK) fnmi Septem- 
t)er i;i to Oetolier 4. Salmon fiy 1.5 inches long were t^ikcn, two or three at a time, 
from Sei>temberl8on. TIk- downstream uugmtiun, thei-efore, Ix'gins at least, as early 
as the middle of September, 

Adult salmon can be found in some part of llie river througlioiit tlie year, and 
the spawning season is therefore very hmg. It is prolmhle that then' are salmon 
spawning at some place in the river or its trilnitjiries in every month of the year. 
They are sjwiwning in considerable numl)ers from July till January, iiielusive. With 
such an extensive spawning pt^riod, it is obviously diRicnlt to separate the young 
according to size, and say that those of a certain size l>eIong to the spring or fall run 
ofa certain year. Avariatiouinrateofgn)Wth, noted elsewhere, adds to the diflieulty. 
However, in the following table of measui'ements of specimens taken with the 
seine at llattle ('recsk fishery during OctolH-r and November, 181^, three sizes may 
be distinguished, which doubtless represent three runs of adults. Those fi-om 
1.4 to 2.2 inches in length wen' from the summer run of IS!»S; the :t,7 Uy 4.7 inch 
specimens from the fall of 1«07 (aiul thoy doubtless wen- among the last to hal^ch); 
and the r..i>-iueh specimens an earlier run, pndmbly the summer run of lH!i7. 

MeHKuri'mriilH »f yoiuiij Miliiiirn liiki'ii villi llir wi'in', BaHir Crri-k, Ortiibt-r IS fii lhi'iniiln-r I, IS'.W. 

1.4 ii 

.1 I 


l.TiiK'hw ... 



a.t IncIwH.... 

sis t«illtii>-bi« 

AZlwlww . .. 


Both f^i'rnj-'iMioii. — An ol>servation station equipi^'d wilii a trap similar to that 
us<k1 at Battle Creek Cisherj' wius established on the river at Halls Kerry, alsjut a 
miles above the mouth of Battle Cn'ek. Oliservatious were made by Mr. (.'hauiber- 
lain, l>eginning .Fanuary >i and closing April ^o, ISO!). The following table gives the 
data obtained at this station. 



Record of salmon fry taken in trap at BalU Ferry, Janiuiry 6 to April 25^ 1899. 







Size, inches 




P.M. 1.4 





' 1.9 ' 2.0 



Jan. 6 

















1 1 
















4 ' 





■ 1 
. . 1 

.! 1.48 






1 .... 



1 ; 


. -_i 



1 14 


10 7 



1 1 


' 1.51 






























1 .... 


...1 i i ■ .... 
1 ' 












i ■ 


1 s 



1 :{l 




'-... i.5i 

' Feb. 1 

.... . - 


1 : 

' 2 




r ' : 







1 "" 








27 ' 
Mar. 1 



: L... 













: i ■ ! 



.... .... .... .... ....1.. _..... 

... '..-. . 

1 . ' 1 






i 1.48 












24 ---- 



1 1 

















_ _ _ . _ 






' 78 


3 1 



H : 



1 .... 


















101 j 



: 1 52 

1 _ 









Apr. 1 

j i i" 




7 1 






1 1 






:::;i:::: ::::;. ;::..::i 


... ...^ .(^ 


1 """ 

. . . . 1 



■ t 

5 i 



















■ ■ 1 




























The preceding table needs hut little explnnation. It indieat^'s that the greater 
part of the yoiiug salmon from tlie fall run paflseil Halls Ferrj" Ix-tween the middle 
of January and tlie middle of March. I*nictienlly all Im^I paflRe<l by March 20. 
Measurements taken January' (>, 13, 21, and ■iO, Fehniarj- 1!), Mareh 7, 15, and 31, and 
April 1, 3, 3, and 4 hIiow that the average size tif those taken on the dates specified, 
tluring a i>erioil of 3A months, varie<l but one-tenth of an inch. The average of all 
measurements is 1.53 inclies. No satisfactory estimaU' of the numlKT {wssing ccmld 
Ik' made, except that therr were pnibably many millions. This n-eonl also pnives 
that salmon fr>' Itegiu migrating as siKin as they an> able t4) swim, and that practi<-ally 
all start downstream at that age, ollicrwise Ihe later ones would liave lM>en lat^r. 

It was also ascertained Ihat a large migration was not coincident with remark- 
ably high water, H is prolwible that when the fry once enter the main river their 
migration is not impeded by low water; but it seems pntbable, fi-om observations 
iiote<1 in another plac*' (see " Sumtner n'sidents") that many of the lat^* fry that 
hatch in the headwaters are detaino<l tliere <luriiig the snmmer by l<»w water. 

Walnut Grnre siniitm. — Our knowledge of migrations thn>ugh the lower part 
of the river was gained froia the genend investigation i»f 1J*!W, and (Specially from 
obsen-ations made by means of a tiiip eslablislxHl at 'Waliiut. (Jrove from .January 
to May, 1899. This trap was constructed especially for the work, but was hardly 
more efficient than traps mtule from seines and used at Hattle (-reek and ilalls Ferr)'. 
It consiste<l of a bag with a short funnel Imng to a 4-fnot hoop, with wings 20 
feet long. At Walnut Hrove the Sacrament" nuikes a sharp bend, <-lmnging its 
direction fn>ui southeast t^j s(>iit>hw4>st. At this tH^nd (ioorgeanna Klougli breaks ofT 
and continues the southeasterly direction of Ihe river above. It thus gets a large 
amount of water, probably half as muoh as the river below, and is in the <linH't 
path of the migrating fry. The ti-ap was set aliout 150 yards fn>ni the heiid of the 
slough, which at that place is about 75 feet wide and 15 to 20 feet deep. The Itanks 
are abrupt and covere<l with bushes. One end of the trap was fastene<l by a long 
roiH! t^o a tree ou the Itaiik, the other to a buoy nuchorcd alKint the middle of the 
stream. It was souu>times set in other positions in the slough or iu the river, but 
without results of [mrtioular value. During a sudden rise in the river it etmld not 
be s(*t on account of tin- great amount of trash in the waU.'r. 

The following gives the record of the cat*-h. In the colu?un headed "A. M." is 
given the number of fry found in the ti-ap at s a. ui., and in the "P. M." column 
the numlsT caught between noon and h p. m. 

Reiitnl iif mttiiumfry taken in tite trap ii^ IViUniif Gnire.Jiiniiari/ T In ifai/ S. /.V.<(». 

1 Cfh. 







]'' i 








ia!».4 aBa.1. M 

1 !:::::: 






'I - 




+ R M 

l.w . 



., -J 



Recont i>f xiilmiia fry tukrn in Ibr Inipot H'alniit Groiv. Jauiiar;/ T In .V'ry A', W)9— Continned. 

*.ai M P.M 1-1 1.-, 1.W1.7 ].» 

1 II 5 . 2 , -! 








1- 1 


SI -:s , at! 


From lliv (ijilii alwve yiveu il will l)v twvn that — 
1. From the mid<lle of January to tlio midUle of May ihenj v 
various numbers passing Walnut Grove, 


2. The height of migration was from March 4 to alxiut the 24tli, about iO days. 

3. On March 8 and 20 tliere were two laryo riiiin of fry, 

4. Practically all had pansed by tiie 22d of April. 

5. The average Mize of tliose Uvkeu duriiig January was inclies, during 
February 1.8 iiichen, during March 1.7 inchcH. From the 30th of March till May 7 
the 8i/« firadually increased from 1.7 inches to 3 inches. 

C. No frj' were talien durinjr tlw first 9 days in January. 

7. Young salmon traveled a^ much during the day as during the nij^lit. 

Comparing the information for Walnut (<n»ve with that for Balls Ferry, as 
shown in the accompanying diagraniK, plate 12, it appears that — 

(ft) Tlie Ballt) Ferry run of February '2 n^achwl Walnut (in»vc Mai-cii S and was 
34 ilays making the dislauce. The fry inci'eaM'd in size n.:( inch. 

{Ii) The Halls Ferry run of Febniarj' 14 reaclicd Walnnt (in>ve March '2it, and 
was 34 days making the distance. The fry increawil in size (1.3 inch. 

{c) The runs that Balls Ferry February 2.1, and later, wen- cauglit by 
high water the latter part of March, which probably carried Ihem downfa.sler. Tlie 
runs were not noticed at Walnut Grove; the net could not Iw workinl during the 
l>eginning of high water, March 23 to 20, during whieh time tliey maj' have [mKsed. 

{'/) The fry taken at Walnut Grove after April 1 hail grown more than 0,3 inch 
since starting downstream, and were therefore the stragglers from the regular 
migration. Those taken during May had prolwibly l»een three montliH on the way. 

It is evident, therefore, that the fry of the regubir migration re<|uire al»out 34 
days t-o pass from Balls Ferry to Walnut Grove. 

The distance l>etweeu the two stations is almut 350 miles. An object floating 
as fast as the current would make the distane/; in almut !) dajs. It requires K days 
for a rise in the river to travel from Ked Bluff to Sacramento. If the frj' travided 
only at night, and simjdy kept with the current, they would make thi> distance in 
18 days. There is no doubt that in migrating the fry drift downstream tail first, 
keeping the head upstream for ease in breathing as well as for eonvciiience in cati'h- 
ing food floating in the water. In this way they would drift much more slowly than 
the current. At Battle (Jreek batehcry fry have l»eeu oliserved traveling with the 
currt^nt, an<l always with the beatl upstream unless frightened. 

The later and larger specimens found had simply been limger ou the way. llie 
larger they became the more slowly they drifted, as they swam against' the current 
more strongly. Those taken at Walnut Grove in January wci-e but I.U inches long, 
being brought down by the high water in January, the short time they had traveled 
being indicated by their smaller size. 

The failure to catch any fry during 1h« fii-st '.* days in January indicates that the 
fry from the summer run had all passed and that those fnim the fall run had not 
yet reached Walmit Gn>ve. Without doubt thcrt^ were a few passing at that lime, 
for there were some passing Battle Creek as late as Decemlnir fl, but tliej' were so few 
that none were taken in the trap. It is possible that there are a few passing down 
the river all summer, though we have been unable to find any after June. 

Ohservatiotui at Beiiicia. — Februarj' 21 and 24 and March 3, IX'.'ll, five siHK-imenH 
1.8 to l.ii inches long were taken in Caniuinez Straits at Benicia. The average size 
j at Walnut Grove after February l'> was l.S inches ami the size of the Bi'nicia s|woi- 

mens indicates a short jmssage between the two places, pi-ol>ably not over a week. 
This would make the time from Battle Cnn^k hatchery to brackish water « weeks. 



Jam ^pri/. 









9ML'.g:i3;j i.i 







■ ■ : ! 















































'Tt't 1 1 1 > 












Safmon. 1 























E I . 






: ■ ■ 





H-> H- . i i , i 







. : : 


tU 1: ^^ 

4l r? J 











ft; \' :;~ 

Si^'* ;■ 



i f 









M ^ 1 -- 

r , ■ 







- t r f - - f r 

ri : i :: 


-T T 



__ . 


4_ t - i- . i i : 


.: ,- 











: > 1: I : 

jznf ; |:: 



-l . '. 









1 : j J 





1 r r ! 1 M 



vrngc length ol llw Iry latea. 






General iin-c.sHijtilion ihirhi'i ISOS. — The followintr note's pive the data obtained 
by the general invest igalion in 1S!IS. 

Tali}e Hliiiiriiiij (jri'tilritt ii'tnibcr "f fn/ tiikxn at oin' haul of the neine. 





iil» 1 


Uii milm b>>l- 















Twuuly niilw bal-iw 

KniitbtH IdDilini! 



if i/outig salmon fukeit in the Sucnimeiito Riv 

From th© tablen given it will \>e seen that— 

1. Young salmon were abundant in the river the first of May, at least Iwtween 
Redding and Tehama. As a few were taken at Sacramento April H'S, A i» probable 
that they «ei-e distributed throughout the river. 

2. The last of May they were nowhere ao abnudHut as they had been at Red 
BIutT and Tehama thi-ee weeks previously, 

'S. A few were found throughout the river May 18 to -tO. 

4. The fry found in tlie river in May had an average size of about -.2 inches. 

5. In July there wei-e no fry found except u few at Keddiug and Battle Creek. 



It in known from I ho work at B»11h Fcrrj- in IS'.'H that pniclicKlIy all ttm fry 
leav» the rivorltefon* March -20. It is evident, tlH'refnri', that the fry fi>nn<l I>etwe«n 
Keddinj! aiul Teliaum in >fay, I^OH, were the Mtra^i^lem left from tlie re^ruhir migra- 
tion. Their size, 3.:J int^Iit-H, iii-eeliiile« tlieir l>eh>iiging lr> the regular run. Tliey 
hatl <'()lleet*<i in Ihe pools whei-e we liid onr wining, as they <lo in tlie liea<lwatt>rs 
(luring tho snmmer (soe notes on "SnunnerresidontM"), wliieh made them apjiearto 
he more Hittindaiit than tliey prolmldy weiv. Tlie rise in the river, whieli (K'<-urn'd 
from May 1-') to I'ii, and (lie a<-<-'>nipanying niuddy water eansei) tliem to jiass down- 
stream, 'i'liis is indieattnl Ity finding fewer in llie pools in Ihe latter part of May 
(when migrating they would not I)0 coUeoled in the ikm>1«) and liy llnding none nt 
all in .Inly. 

The CKinditions in iHUH wen' exceptional un account of the early m-currenee of low 
water— from the middle of March till the middle of May — though (lonhtlesstheiv ar« 
always a few stragglers fnim the ivgular migrations. Tliese deen-ase in number 
and iucivasti in size (sliglitly) till the rains of the following winter, wlicn all leave 
t.ho river. 

In the nioiinlHin .streams the young salmon prefer the ikmiIs, where they are 

ne in a jiool at thn 
111 at all uneommou 
I headwaters. The 

were wean*ely ever 

often abundant. Nearly aOi) wen? tnken at <inc haul of the 
head of Ko.\ Canyon, near Sisson, in August, 181i7, and it wa: 
to catch over a huuilred at a time in many of tlie pools of 
i:api<ls have l>eeii fished a nnmlK>r of times, but. young salm 
caught unless the watt'i" was at hnist il feet lieep. 

lielow Uedding min-e salmon weiv foiiml in the wat<'i' with mothTate curn'nt, 
gravelly br>ttoiii, and a depth of over iM feet, but none was found in al)Soiutely still 
waiter, anil none over a soft mud bott.oni. Not inueh seining was done over r<H'ky 
Ixittoni, on aoeonnt tif the stroag etiri-eiit and the injury to the seine liy its picking 
up eobltlest lines. A few satni<iii wen- eauglil by putting euougli lloats on the seine 
to keep it at the surface anil then hauling in water 15 feet. deep. 

The foUowing table indicates the various characters of stri'am in wliieh young 
salmon were found in (be main river, with the numlx-r taken in one haul of the Koine 
under the various iiinditions. 

T,il,h' nhiiwiHij V 


Ht-ildlni: Mil 

M'lDllM'f linitli riT.'k 

Bntt«<«y I 

TtiiinllH.lifl.'wl'rinii'l.Hi . 

i mulvK.'."' ".;.:;. !;;!: 

I KniBhtH _, 

Swrnoonfi ^ A|ir! 

I KluVtotii May 






Morfin*-nfs in f.sfunrtj nnd hdij. — Mii<»li seining was done both years at Oleiiia 
in Irvinu t(» learn sonietliiiii^ ot* the ni(»veinents <»f vouu;^ siilmon in brackish water. 
None was found in is*.»,s. A few were caught near the mouth of Paper-mill Creek in 
1807, and one was taken - or '-) miles from the mouth of the creek, across the liead 
of the bav. The net was stretched across the mouth of the <'reck for 15 minutes 
during the flotxl-tide, and two salmon fry w(M-e taken, indicating that they run ba<*k 
and forth with the tide. 

The iishermen ai Marshall, on Tomales r>ay, about l^> miles from th(» mouth of 
Paper-mill Creek, reported having taken young salmon inconsiderable numbers the 
last of April, 181»7, about 50 days aft^M* they were liberated in the streams near 
Olema. At that time the salmon were alxjut ICX) days old and were large enough to 
be taken in the seines used by tlu* fisliermcn. Tliey caught as many as 15 or iM) at 
a haul for about a week, and caught them occasionally till the middle of June. \ 
think the report reliable, as tlie salmon was a new fish for the bay, and would 
attract much attention. This indicates that the fry may reach the ocean at the age 
of three montlis. The water at Marshall is pure sea water. 

Specimens liave been taken in brackish water in Suisun and San Pablo bays, 
but not enough to determine their movements. A f(»w about 10 weeks ohl have been 
taken at Beni(*ia in water that was about -0 i)er cent sea water. 

Effirt of sea wafer on alevins and fry, — To determine the effect of sea water (m 
alevins and fry, i25,(>(Mj eyed eggs were taken from Battle Creek liatchery to t,he 
Hopkins S4»asid(» Laboratory at Pacific (irove. The eggs were received at the labora- 
tory December lo, ISOS, and most of tliem liatched on the 17th, which date was taken 
as tlie basis for determining their age in the various experiments. Those not being 
experimented with were cared f )r as alevins ami fry ordinarily nn\ 

The first exi>eriments were made by putting a few alevins directly from fresh 
water into batt<*ry jars filled with various mixtures of fresh and sea water. In the 
later experiments glass tanks 2 and 3 feet long were used, and the water was kept 
running. The experiments were begun when th<» alevins were <> days old. It was 
found that at this age they could live in<lefinitely in water that was 25 per cent sea 
water. Those about 40 days old could live in 50 [nn* ct»nt sea water, and at 50 days 
75 per <*ent. Those GO days old couhl live in l»5 i)er cent, though there was consid- 
eni])le h>ss. Ninety-five percent was as nearly pure sea water as could be obtained, 
tlie lalxjratory pump being broken and the tank partly filled with fi-esli water. The 
loss was much less when the density alternate<l between a high and low i)ereentage, 
which indicates the value of the change of density in the estuaries with the rise and 
fall of the tides. 

When the younger alevins were placed in 5o per cent sea water or stronger, 

the yolk was solidified, lM*coming much like soft rublx»r. The blrnxl was driven 

from the body, making it. appear bleached, and th<» adipose membrant* at each (»dge 

of the tail a<lja<*ent to the caudal fin turne<l white. Th<» circulation was retarded 

an<l the fish l)ecame sluggish. The only noticeable effects on the old<»r alevins were 

sluggish movements and an inability to keep a horizontal position. Sometimes 

death was immediately prece<led by violent and spasmodic swimming in any and all 

directions. The Siime actions were noticed in minnows placed in a sti-ong mixture 

of sea water. 

¥. c. B. rjou— 7 




Tlio following table giv<»s h nHMml of the oxiH^riiiients. The iKM^MMitajri* of sea 
water is reekoiKMl ]>y taking piiiv sea water as a staiidaixl of 100, a iiiixtiin» of equal 
parts of fivsli an<l sea water being 50 i)er cent. Tlie pereentage of <lea<l ah^vins hI any 
one time is not ])as(Hl on tin* original nunilM»r with whi(*htheexi)erini(Mit Iw^gan, but on 
tlu» number left in th<» vesscd the previous day, e, </., if we start with 40 alevins and 
10 die the first day, that is 25 per eent ; if 15 die tlu» seeoinl day, that is 50 \h}v cent. 

Tnhlv t<howhig effvvt of nalt initcr tm nlvvitiM. 

i\ ■ 





Lhiy of 




Day of 












































































IVrrvntujrt" died. 

Exp. 1, Exp. 2. 


25 iK»r 

r«*nt wa 


(I 2 








(» o 

2 17 



(I • 8 








Exp. 8. Exp. 4. Exp. 5. Exp. fi. 

5H) iK»r 


i.> iK*r 

<*fnt m-H 






Si« water, fnmi No. 





Exp. 7. 



from N<». 




lH«nnity raiHe<i ^radunlly to 

Per<'entajf«» <lio<l. 


I * 


i Exp. 8. 


Exp. 0. 











into Heu 


Exp. 10. Exp. II. 


ExiN'rinient 12. 
, Pereent- | 

H)fe HKHl 


aK«* di<><l. 

,'»/M'r«-» n/.| 














































;i ; 







1 : ! 




Table tfhoicing t'ffect of sit It tntter on alevius ami fry. 

-1^- - — 

Day of 












age sea 
















. > atroHett 











. 5l> 



ajre si-a 

















51 > 







































s ; 




tinue<i. ! 

ExiKM-iiiuMit 1 was iua<l<» to clieck th<» others, showing that the h)ss was not the 
result of being i*ontine<i in battery jars. 

ExperiniiMit - sliows that 25 ^hm* <'ent sea water has but littU* (h*k»terious effect 
uiK)n alevins over 5 days old. 

ExiM»rinieiils o to 7 show that ahnins of to 10 days of age can not live in sea 
water of oM per cent or over, c»ither when put directly from fix*sh water into tlie 
mixture or when tin* deiisilv is graduallv raised. 

ExiK*rinients S to 1 1 show that alevins 12 days old can live longer in 5n and 75 
per cent sea water than those only (3 days old. Some of the 12-day alevins liveil 
7 (lavs in 5o iM»r cent s<»a wat<*r, while the ^J-dav ones lived but 3 davs. Tlie older 
alevins lived one day longei* in 75 per c<*nt, but died the first day in pui-e sea water, 
as did the 0-day individuals. 

Exi>eriment 12 shows that when TJ days old they live longer yet in 50 per cent. 
Two lived 2 days in 75 jH'r cent aft<»r having been in 50 i)er cent for U days. The 
one that lived till the twelfth day was a week older. 

Experiment 13 shows that a gmdual rise from a density of 10 iH*r cent sea water 
when 20 days old to 75 i>er cent when 32 days old was fatal. In a similar way a 
gradual rise fnmi fresli water wlien 33 days old to a density of *M) per cent when 44 
days old was fatal, as is shown in experiment 15. 

Experiment 10 indicates that alevins 50 to GO days old can Ix^ar a high density 
of S4»a water, but that thev can withstand it better if the densitv, instead of increas- 
ing n^gularly, alternates Ix^tween high and low. It does not show* the exact age at 
which the fry can live in sea water, an<l it is <loublful whether this can be determine<l 
accurately in aquaria. Whether it can or not, the time wjis not at my disi^osal to 
carrv the observations further. 

As a wholes the experiments show three important i)oints. First, the fry can 
not live in sea water until Si»veral weeks aft<*r the volk-sac is absorbed; second, when 


able t» live in 8ea walor tliey i-hii not. no ilirfctly rivmi Tivsli wtiicr to wa water, bat 
iiiiist paJWKt-ailiiMlly; third, Uicyarecn-ally aidt-d by nii altcniali<iii ordi'iisiliosHiich 
as is obtained by iiaNsin^ ibi-oii^li aiii-stuary. I-'m- tb«'.se ifasoiis il woul<l imt bo wtsll 
to plant fry in aKti-cam tlitil dmts not ivai'li tli<> oi-^-aii Ilin>n^li an est nary. 

Vhaniji- itf ailor tliirlnij miijriif imt, —Tin- f<i\or tif yiMitu salnion dt'iH-nds much 
on till' chamet*-!' of llii' watJ-r in whicli tlicy live. 'I'hosi' in small, cold slrcaiiitt an' 
much more duaky and liavo Ihi' parr marks Kironfjly <lcvelo|>i^<l. 'I'lu-y iHicomi! 
UghtiT in color upon ciiliTiiifi tin- main river. Those in braekisli and sail wabrrare 
bright silvery on the sides, with Ihe back .sea pven. I'arrs 4 t<i H inches in bnif^th, 
found in Hattlo Creek anil similar pbuvs, have the siili-s brijrht. silv<'ry, tlie btwk 
olive brown, with the upi>crcnd of thii parr marks niakiiij; ii'gnlar sliadin^'K along 
the biu;k. Specimens 'J.d incln's in lenglh from 11<m1i-<>, San I'ahlii liay. May is, have 
distinct parr marks; 'i.7 inches, fi-um ISenicia, May i:t, have liiNt them. Sometimes 
the caudal lin is reildish; sometimes there arc yellowish strijH's on the vent ml ami 
anal fins, ti»peeiHlly with fryNilH)ut l.-i inches lon^. 

finnntuirij of ahxirvnthnu on mi;/i''ilion.— The fry iM'fiin their downsi ream niigni- 
tion as soon as they an- able to swim. In the clear water they travel nioi-e at night; 
in muddy water, as mnch or iiioiv during the day. Much of the time they Hoat 
downstream tail first, anil in Ihe larj;er streams they travel moi-e or less in schools. 
In the larger streams lhi'ir<hiwnslrcam movement is not dependent npim the height 
of the water, but upon age. Fii>ni OcIoImm- to April, hielnsive, over H!i i>er cent that 
liass the vicinity of Battle (.^leek are of the .same si/.)-, inches long. They pass 
down the river at the rate of about U' miles a day, and are almut ti weeks reaching 
brickish wat4-r, l>eing ;i months old at that time. They are proliai>ly 4 or 5 months 
old when they reach the ocean. The ebb and How of the tide in Ihe estuary, causing 
an altonnitioii in the density of the water, is apimrenlly heneftcial. 


Geiieriil. ocroinil. — In the upjUT i>ortion of the Sueramenio Uiver Ihere yet 
ramained, after the winter an<l spring nii<>ration in I.S'.IK, a large numlx-rof young 
Halmuu. In the vicinity of Sims we found fnim 70(i lo i.iHKJ in the various {xtols. 
We found tliem eommou in the ^leCloud at Itainl in September, ami in Fall River 
in AugUHl. These slimmer ntsidenls, as they may be called, are confimKi to the 
headwaters — the clear stivams with i-ocky Imttums. They do not stay much of the 
time in the Very swift current or riffles, but remain in the muii' quiei [kmiIs. Here 
they fe(!<l on aquatic insects and take the angler's lly tlic same as trout. 

Moat of the data concerning Ihe summer residents was olitaincd fi-om investi- 
gations near Sims, in Hazel Creek, and the river l«dow its mouth. Hazel Creek is 
a small mountain stream, with many i>ools and gravelly riffles, and is a favorite 
spawning stream Ixith for Halnion ami trout. The two lower pools, which are about 
a (juarter of a mile from the mouth, were .seined several times during Ihe summer 
and fall, and itwjw from this work that we learned mneli that wc know of the habits 
of the fry i-emaining in the streams during the summer. In the table iHdow one of 
these ]K>uls is called the u])]>er and Ihe otiier Ihe lower. 

The Sacramento lliver in the vicinity of Sims is alwint 4ii or 50 feet wide, and 
during the .-^ummer has an aversigt; deplli of about '.i feel. It is very swift except 

V ■.■w.-h other tu bv tliu n 


in tlie pools, which were the only placen that could be seined. Seven of these pools 
were seined frequently, and for oonvenieuee in keeping notes we numbered them 
A, B, C, D, E, F, and G, Ix^ginuing with the upi>er. 

In July and August all s^)eeimens taken in Hazel Creek and in the river near 
by were marked by cutting oflf the adipose fin with a pair of small curved scissoi-s. 
This enabled us to know when we were taking specimens that had been taken before. 

Upsfream movenieni, — The following is a record of the seining in Hazel Creek. 
The dat^i for each seine haul consists of the date, the pool where it was made, the 
number of fry previously marked in this pool, numl>er of days since the last were 
marked, total number taken, and number of marked fishes taken. 

Record of seine hauls ht Huzel Creek. 

! Previ- Days Total ^'^tohof 

Date. Pool. onsly sin«*e last ,.j.*:^ marked 

marke<i. marking. fish. 

May 17. Ix)wer U 

July 9 do .55 

Aug. 17 do. 4» :» 49 \2 .; 4^ :W 45 5 

! I 4^ :*♦ :«i ' 2 

I Upper. I 7M , 1 

'' 5.3 . 1 37 I 





Catch of 





since last 



LowiT . 









Lower . 










Nov. IS. 

Lower . 










I>«-. IS.. 

L<jwer . 





L'p|>er . 




1 U 



It will be seen from the ta])le that in seininj^ the lower ik>o1 in Au^rust, V.i young 
salmon were obtaine<l that had been marked in Jul v. As there had been but 48 
marked in this pool in July, it indicates that at least 40 per cent of the fishes that 
were in this pool July remaine<l until August 17. 

In the upper pool we found two July-marked fishes in Aufi:ust, where none had 
been marked in July. As the lower pool is al)out KK) yards from the upper, tliis 
indicates that at least 4 per cent of the fishes in the lower pool had ascended the 
stream that distance*. 

There were 104 siMM'imeiis marked in the lower pool diirinj^ July an<l August; 
31>, or 38 per cent of these*, were* found there in Septeml)er; 3»), or '.I't iM»r cent of the 
107 marked in the upper pool in August, wen* found in SeptemlnM*. Only <me hfiul 
of the seine was made in each j)lac<' in SeptemlM»r. 

In August four-sevenths of th<» marked fish(*s found in the low(»r pool were taken 
in the first haul of the seine. (When more than one haul was made the marked fishes 
taken were hehl till the seining was over, in order that they might not l)e cemnted 
twice.) Assuming tliat Xhc same x)rojM)rti<m was taken in the one haul in September, 
we would reason that there were <>8 nutrke<l salmon in the lower pool that month; 68 
would be ()5 i>er <*ent of the numlx^r marke<l — that is, 05 per cent of the fishes in Hazel 
Creek on August 17 i*t»mained until Sej)tember is. This api)roximates the (\stimates 
made for the pools in the river below the mouth of Hazel Creek. (See notes Indow.) 

In a similar way, 'M per cent of the 104 marked in th<* lower p<M)l were found in 
one haul in ()ctolH»r, an<l 45 i>er (»ent of the 1<'7 marked in the upi)er p<K)l. Thus 
there was a loss of i* per cent over the previous m«)nth in the lower j)ool, and a gain 
of 10 per cent in the upper, in<licating an upwanl movement. 

But the upward movement is indicated l)etter by the simph* statement of num- 
bers, as given in the table. In Septeml>er there were 39 marked fishes taken in the 
lower pool to 30 in the ui)per; in Octol)er tin* ratio was 37 to 48, and in November it 
was 22 to 54. It is difficult to see how this can mean anything else than that the 


yonn^ shIiuoii in Hazel Crook eoiitiniied to work thoir way upstn'aiii diiriii^ Sop- 
UmuIxm*, Octohor, and Novoinl)er. 

This table shows an increase in tlu» nnnilxM'of salmon in Oetober. Tin* average 
of two hanls in May was 17; one in Jnly pivf 55; six in Anji:ust ifavc» an av<»ra^e of 
4*J; two in September avei'aji:(4l 4<); two in ()<'1ob<»rlin; two in XovtMnlMM* l<**i. The 
simple fact of there IxMnj^ more fishes found on hiter <lat«\s would n<»t indieate an 
upwanl mo\ement; they mi^ht have como downstream. The larj^er ])ercenta^e of 
marked fishes in the upiH»r pool, however, would indiratcan upstns'im niov<*m<Mit. 

Several plaees alM)ve the uppt»r pool wen» fished in <><'tolM»r — one plact* within l/» 
yards of it — but no marked sjilmon wen* foun<l. This would indieate a lark of upwaitl 
movenuMit ; but, all the <lata Immu^ <*onsidered, it is ovidrni that lln^n* was at least 
a slij^ht upward movement in ()<*to]>er and Nov(MnlM»r. 

In I)eeeml>er the you n^ salmon had all disappran^d from th<* u^^mm* pool, and only 
4 wer<» found in the h)wer. One of these was a si)eeim<Mi mark<Ml in July or August. 
That is, I out of 2<K) remained after the DeeemlH^r ris<». It is <'vid<Mit from this that 
praetieally all the youn^ salmcui left thoen^ek bi'lw(»en NovemlM*r IS and I)eeemlK>r 
IS. There was a heavy rain in the vicinity on Nov«»mb(»r l*s an<l iM» (preeipitntion 
over '2 inches), and then* can be little doubt that the salmon all h*ft at that tinu». 

Miijrat'hm durunj siunnif /*. — Pool H (se(» plate 1 1 ) of the riv<»r is srparati'd frcuii 
I Pool A (at the mouth of llazrl C-ivek) by a rapid al)out 150 yanls lonjx with a fall of 

about 4 feet. It is over feet deep, with larjje angular rocks alon^ one <m1«j:(* afford - 
^ in^ excellent hid inji:- places for y<mng salmon. It nM|uin*<l thnM» men to soine this 

[ jKKd well, one to throw out the seine from a lar^jje roc^k at the iippor t»nd, and two to 

I pull it in. When there wen> but two of us, one would jwiy out the s«Mn(» fmm a 

riffle alx>ve the po(d; the other would wade out as far as possible in tin* lower end 
of the iK)ol aiul pull the seine down with a roj)e. When tin* s(Mn(» was stretched 
through the pool it was pulle<l ashore. Obviously such work was not very sat isfactory. 

The followinji; table <j;iv<»s a reconl of the catch at each haul of Ww s(»int* in P<»ol 
B in Jul}' and Auji^ust. Four hauls were nmde in this pool in May, alniut 50 young 
salmon bein;^ taken in erne haul. The specimc^ns caught in May n^[)n'sented two 
sizes, su<»h as wore found in Hazel Civekon the same date, but thon* were very f(»w of 
tlu? smaller size. Tlu»se smallest were probably from a few late-s]>awnin<r individuals. 

The table ^ives for each haul of the* seine in July and Au«i:us1, (1) tho dat^», (:?) 
numlMM' of young salmon (viught, (-)) number previously marked and n'lcas<»d in the 
pool in July, (4) tim<» sinc(» the July-mark«Ml fishes wen* released, (5) numlN^r of July- 
marked tish(»s caught, ((3) number previously marked in August, (7) time sin<*e the 
August -nuirk(»d fishes w<M'e relea.sed, (S) numlK^r of August -marked fishes taken. 

These data may be used to estimate the nundier of young salmon in the pool by 
making th<» following proporti<ni for any particuhir haul of \hr seine: The number 
of marked fishes taken is to the total numlK»r of nuirked fishes known to Im^ in the 
f • iM)ol (having just l)een released), as the total number taken is to fhf fnfal numln'r 

til the pool. The results, of coui'se, art* variabh*, and it is only by a number of trials 
that we can get near the probable truth. Th<» value of tht* estimates is not enhanc(»d 
by then* being no marked fishes f^ken at cc^rtain hauls. In sueh cases, how<»ver, 
then^ were but few of either kind. The <^stimato made fn)m ea<*h seine haul is given 
in th<^ table. 

The fi-eshly marked siK^cimens could l>o distinguished from those marked a 
nu)nth previously by the latter having the scar healed. 





Table of data obtained from seining Pool B. 


Num- Number 

berof prt'vi- 

youDj? ously 

'salmon inarkinl 

k'anght in July. 

Time sint'H 
July mark- 


f»f July- 




uiarke<l in 

Time since 
last AujJTU-st- 

fisht'H were 


Numljer of Estimated 
August- number 


of voung 


in iMM)l. 

July 9 . 

Aug. 14. 
Auf^. 15. 

; Aupr. 16- 

: Aug. IS.... 













1 to 2 hours. 

3 hours 

J* <lays 

;I7 days 



44J days 


















1 to 4 hours. 1 

, 1 day 1 

I 3i» minutes .1 o 

...-do ' 1 ■ 

1 diiy I 

4 hours j 9 

31J minutes.' 3 

2davs 1 

435 ' 

57S , 

1,595 I 


42:^ i 


Th(» eliu»f value of this table is in the estimates <^iveii in the hist column. The 
estimate of the numl)er of young salmon in tlie pool rang<\s from 415 to l,r)l»5. Half 
of the estimates come within 107 of the average, which is OSo. This average is prob- 
ably not far from the actual numlM»r. 

Pool C is qiiit^ similar to Pool B, and was seined in much the same way. When 
two worke<l it, one had to hang the net on the rocks <m one side, then swim across 
to the other side, when botli pulled the seine off the rocks and haule<l it inshore at 
the lower end of the i>ool. 

Tlie following table gives a record for Pool C similar to that given for Pool B. 
In the only haul made in July, two of the four fislies taken had been marked. 
They had evidently come down from the pool above. Likewise in August four 
August-mark(*d fishes were taken in thefii'st haul, though none had yet l>een marked 
in this pool. In making the estimates of the number in the pool these four are 
considered, being added to the '*numlxM* previously niarke<l in August.'" 

Table of seine hauls in Pool C. 


1 Number . 
of young 
I salmr»n 
I taken. 

Xumb»»r ' 



of July- 

in July. ,"^rkmg. ^^„^^^ 

July 9 

August 15. 

nmrked in 

Time .si nee 
last mark- 

Numlier f)f 
A u gust- 
mar kiKl 



numlx'r in 

pix)l in 


August Kt . . 
August IH... 





36 days 











?n davs. 





39 days. 






2 hours 

31) minutes. 

1 day 

S) minutes . 

2 days 




1,91K 1 



:M4 I 

566 , 


The data obtained from Pool C gives a larger estimate for the total number 
in the ik)o1 than that for Pool B (C, 801; B, 085). In the seven hauls of the seine in 
this pool in August there were S July-marked fishes secured. In the t^n hauls in 
Pool B VJ were secured. From these two statements we determine that 18 per 
cent of the July-marked fishes of the two pools were in the lower iX)ol in August. 
(8-7-7x10=11, 41>-f ll = GO, ]l-f-6()=.18.) Only 1 percent was released there, leaving 
17 per cent to migrate. Som<» of these may have drifted over while faint from being 
confined in the net, but we think not many. We never saw any do so, though we 
often watched them for that purpose. It is safe to say that most of them went 




over voUint'»rily. Wo would expect this, as the connection Iwlwccii the two pools 
is quite deep, thoufjli swift. It is oven roinai-k;il>lc that. ii<) iiion' Ilian 17 percent 
passed to the lowei- pool. It is (.'oncwlcd thai the estimate's in the tallies al)ove are 
liable t.o considerahle errors. There are always some unknown quantities in the 
equations, yet the results appear trustworthy. 

Pool D is below the lower railroad liridRP, anil was quite unimportant. It was 
seined but onc^v a lai'i^e rock at the lower end of the pool makin;; seining inipraeti- 
eablo. Six young salmon were taken. It was one (tf these six that wjis tiiken in 
the first haul in Pool E. The two pools are eontiniious. 

Pool E is a iH)rtion of the river about 75 yaiils Iouk, immediately below ami not 
separated fnmi I), ending above a long riffle. It was the only piHil that could lip 
entiivly eoven-d by the B<Mne. The bottom is mostly covered with cobblestones, and 
there are larfre mcks alonj^ one shore. These ufFoi-ded hiiliiig-plaees for the younj( 
salmon wliile the seine was Iwing drawn. It was seined many limes in August. 

The following table gives the record of the cateh at eiM-h haul of the seine niatle 
in the pool, giving (1) the dale, (;;) the iinudn'r caught, (;!) the number previously 
marked and released in the ixiol, (4) the number of markcil fishes taken, (5) the 
estimate of the nnnilx-r of young salmon in the pool, and (<>) the variation of thin 
estimat.1^ from the averag*.- estinmte. 4>a the Hilli and 17th of August the seining 
was earned <ni continuously, the time retpiired for making a haul and counting and 
marking the parrs being fi-oin iio \<i ;(i' minutes. The nmrkinl speeinn'u Ijiken hi the 
first haul was one of the six from Pool D, 

liirorit <•/ xfhu- hmih in i-ool K. 

j .-uuBlit. I 


'llie purpose of llie work in tills pool was t,> d<-1enninc tlic iinndH-r of young 
salmon thai might Iw found in a pool. The cslimatcs arc maile in llic same way as 
in the cmte of Pools It aiul ('. The Iliii-<1 liaul, as noted in the table, may Im> taken 
as an example: :til, the innnl>erof marke<l tishes taken, is to I ',i7, 1 lie ii u ml hm- marked 
previous to this haul, as U(i, the total nundH-r taken, is Id u:i.\ tl,, l„f,tl ninnlwr in 
Ihf imil. The estinmtcs vary from .5.S(i to l,8i>l), bnt .si'vcmi of I hem an- not far from 
a Ihousjmd. The avenigeof the estinmtea isl,o:>J, which is pi-olutbly near I he I ruth. 

S<;iitcml)cr IK we seined the i>ool )M;ain, catching 17S young sjdmoii, '.'7 of which, 
a tittle over half, had Im-cii marked. l!y looking at the table above it can l«' seen 
thata little over half of the estimated numlM'r in the |mio1 were maiki'tl in .\ngusl — 
o8l) out of l,Ui;i. If an estimate U made of the number in the pool in SeptemlR-r, hy 


assumiu^ that all the marked fishes remained in the pool, the result will be the fol- 
lowing proportion: 07, the number of marked fishes taken in Septeml>er, is to 580, 
the number marked during August, as 17S, the total catch in September, is to 1,075, 
the total number in the poo!^ which is remarkably close to 1,0:22. 

If there had been much of a migration between August 18 and September 18 we 
would have taken a larger proportion of unmarked fishes. If there were no migration 
we wouhl expect to get marked fishes in proi)ortion to the total catch as 586 to l,02i\ 
which was the proportion of marked fishes in the pool in August. The record for 
Septemlx'r, 97 marked out of a total of 178, is very nearly that ratio, indicating 
that there was little or no migration l)etween August 18 and September 18. 

August 17 we fished two pools, F and G, alx)ut a quarter and a half mile, 
resx)ectively, l>elow where any fish had been marked, catching 13 and 115 salmon. 
We also caught 57 in Pool (i in Septemlx^r. Neither at this nor at any other time 
have we found markcMl salmon below the pool in which they were released, except 
in the case of Pool C, above referred to. The one marked salmon taken in the first 
haul in Pool E was reh^ased in the upper (»nd of that pool. 

It is especially worthy of note that none of the 591 fishes taken in Pool E had 
been markcnl in the pools alK>ve the previous month. If any had left the upper 
pools, they had not stopped in the vicinity. 

Sfufimarf/ on nnniher ami movenifnts. — The estimate of the number of young 
salmon in August in Pool B is ti85, in C 801, and in E 1,022. Pool E is much longer 
than the othei's, and might very well have more fishes than either. From these 
estimates it is probable that there were al>out 10,o<M) young salmon to the mile in 
the Upper Sacramento during the summer of 1898, or between a half and three- 
quarters of a million in all the headwaters of tliat stream. 

There is little migration of the young salmon l>etween May and December. 
Where pools are separated by shallow rifHes, no evidences of migration could be 
found. If connected by deep water, it was found that alK)Ut 17 p4»r cent of those in 
the u[)per passed to the lower. In Hazel Creek there was an upward migration of 
4 iH»r cent during July, and a larg(»r — alwuit 12 i)er cent — during Septeml)er. There 
was no diminution in number either in Hazel Creek or the river up to November — 
even an apparent incivase. There was a sliglit migration, however, during the 
whole period, indi<'ated by the disappearance of the larger marke<l siH»cimens. 

This residence in tlie headwaters during tlie summer is [)robably due to low 
water. It has Ixmmi noticed many times, lM)th in the streams and in the hatcheries, 
that young s^ilmon <lread going over a fall. When the riv(»r is very low, as it is 
every summer, the ra[)ids In^come almost like waterfalls, thus preventing down- 
stream migration. A slight rise obliterates the fall and at the same time makes it 
difficult to find food; hence the decided migration in l)eceml>er. The abundance of 
food appeal's to be of some importance when we notice that there was a scarcity 
of food in September and also a slight increase in migration that month, although 
the water was the lowest of the season. 

Young salmon were reix)rte<l abundant in the fK)ols near Sims the 1st of May, 
1.S99, and they doubtless remained during the summer, as was found during 1808. 

Grou'tli In fre^sh water. — It has been shown alM)ve ("Migration, general investi- 
gations of 1898") that the size of young salmon found in May was the same for all 
parts of the river. This was true also in July for the portions of the river in which 
they were found. 



The following hIimws the moas tire men ts of young snlnioii ttikeii hi tl»' Siicmmento 
iuJiily, ISitS. Hol<i-fiM'cd typo iiulicati'.sw-)ii'n? the HvcnvgeHizi's full. FDrDiniHumir, 
Keililiiii;, Hiiil ]inttl«>OiH>t'k Ihir niiinlK>rHgiv«>ii iiKlicati* the total iiiimlHTof lish liiken. 
NoiK' iveii' Ijikou lielow the Utter puiiit. All weit> tiikeii iK-tweeii July it and K!, 

TViWc ubinriiiy xUe of jiiiiiug Milnuiii tiikcii ilnring July. ;,v,'W. 

Nuuilr-r iJ HiuH'ilii.'iM. NiiinlHT i>t i>|>»'Iiiu>dh. 

"'"■- Dtin»- : ^:,„ IlHii-l H.1I- IHtUu *""■■ , Duiw- , .,,„„ ■ Ha»l Rt^d ' B»ttl.> 1 

IJJ - 


Note that thi'iivon^i;i'si)ti'.s in tlittalxtvo tiiblo iirofnini H.H inehcs in I[»/a-I ('reck 
to 3.1 inches at Iteildin;;. In i-nnijuiring this tiihie with that for the month of May 
(see iilwvc), it will Ik- swn that tlu' averuge size increased fnnn 2.4 inches In Mhj' to 
3 inches in -Inly, an increase of O.ij inch in two months. This is also the amount 
of inci-eaw if only the smallest siMHUinens in lla/.cl ('iM-ek are considered. They 
incn-asetl from i.J' inches in .May t<i 2.1 inches in July. This is an inen-ase of 0.3 
inch iMT month for lishes averafjing under :i inches ia length. Knch tahle shows 
that there was a greater variation at the upiior stations. 

The growth of ",-t iiieh per month is also shown by the following tahle of meas- 
urcmenis of specimens laken at Kissou in May and Angust. The average sixes are 
indicated hy heiivy-fa<'i'il type. In computing tli<- nvcrjige for May, (he four largCHt 
lish are not conn ted, as they evidently l>elonged to the summer run of adults iuslt^ad 
of the fall. Measurements of si>ecimens taken in the river at the month of the cn>ck 
in Angust an* given for eouiparisou wilh those fnnn the civek. The average size 
in .May was 2.2 inches; in August, a little over three months later, it was :i.:t inches. 

tile incre-iksf^ in si/e of those remaining in 
The growth was pnit)ahly a little gn-ater 

Tiililf Klioiriiii/ iiwrriiKf ill k. 


k Itetng one-lhini inch in-r month, 
thai, the larger s^H-eimcnH migrating! 

■ 'if yimiiij mlJniiin iit Sim 

■ Sultaw«y»T.^.k-| 



- ' 1 * 


The above .■<huws the amount of variation in the young salmon of apprii.\iimitely 
the same age. All were released from the Sisuon hatchery. The ohlest were hatched 



Det-enOxT :?3, 18i»7, ami the youiigri'St January 23, 1898. The hii^est wei-e 2.0 inches 
lou^, thf smallest l.«;, a tlitttM-eneeof 1.3 inches, which can not lie accnunted for by the 
one niontli diffeifnce in a^c. Then* was still a difference oi 1 inch in Augnst, when 
they weiv 'i.S anil i*.s inches lonjx- 

Tile folKiwinjx is a table «»f niea:snro mentis of specimens taken in Hazel Cre«*k 
during vari<ms niitntlis. They aiv thoutrhl to Ik- n»piv>entalive, though th«»siH*cimens 
wei-e seh^etoil. \V»» pirketl o»it extremes an<l what w«» thon«rht to Ix' average sizes. 


May. .Tilly Si-pT. !>»■••. 


Num)*-r iti«-:iMir«<«1. 
May. .Iiily. S^-jit. I>.-* 

l.'t iiii*1i»*s . 






... 1 


l.ti iu'-lu-^ . . . 

1.7 ini'li'-s 


i.l iin-h*"?* 

2M in'* Ill's 



1 ; 

:<.l imhes 

. .. . s.iiiiihtv 

'<' inrhi.'S' 



4.:f in«b»-> 

t :<inrln-i 



2.rt i!i»-h«>< . . 
S.7 in«'h»^ . . . . 
2.** im-lit-^ . . . . 




TliealM»v«» measun*mi*nts indicate* twoages in May, but the yi»ungesl weiv doubt- 
leas fnnn a few, pr«>l»ably a singh* pair i»f tislies. that spawned mueh later than usual. 
The iddest wen* from the regular fall run of adults. The dilToi'cnce In^tween these 
two sizes in >Iav was ".■♦ inch. The two sizt»s are not discernible after Mav, those 
shown in the table Ixing dur ti> the seliH*tions i»f specimens, which is not the case 
for Mav, liowever. Tin* growth is indicat^nl bv the increase in .'^ize of the smaller 
specimens. The smaUest specimens were: In May 1.-5 inches, July 2.1 inches, Sei>- 
teml>er l\s inches, D»*cemlK*r 3.1 inches, the intervening iKM'io<l in ea«'h case Ijeing 
2, 2j aud 3 m(»nth.s, and thr increase* l>eing o.i;, n.7, and <».3 inch, ivsiK^ctively, or 
0.30, u..3.\ and ".10 inch ikm* month. The total gi*owth in 7 months, as shown by the 
smallest siKM-imens. was only l.'l inches, and for the hist 3 months (». 1 per month. 

PiMd A is at the mouth of Hazel Creek It is a si*micin*ular ihx»1 i>f quiet water 
at one siih* of, but not al all separated from, tin* main chann**!. It is over ♦'• feet 
deep, antl lh»» st*ine hatl ti> !>«» hauled by means of roi^e.^. As tin* s«»ine was stretched 
aci-oss the mouth of the jmioI and liaul«*il in at the upiK^r iMnl, with the cuds close to 
the banks, thtMt* was bin little chamt* ff»rth«* t<i es**aiH*. The |iof»l was tished 
nionthlv, iM'irinning with Auirust. 

The follnwing table gives the number and size o( the young salmon taken in 
Pool A during the seasc»n, and alsf» indicates how many wimv marked lishes and had 
therefore remained sinc«» August. As tin* measurements weiv made on live li.shes 
thev could not be ma<le accuratelv enough to be given in tenths. Then* was one 
2-inch lish taken in OcloU'r. but it was not counted, as it evidentlv lx*Uinjr«'d to a 
different run. We marked and retiirinnl to the pool all of the tlslies taken in 
.Vugiist. None wen* mark<Hl in any other month. At anoiluM- haul in SeptrmlH.*r, 
not n^cordtMl in the table, >i2 specimens wnc tak<»n, 32 bt'iiig marke«l. whicli was a 
largi*r prop<)rtion by 7 per cent than in thf haul reci»rdc<l in the tabh» fi»r that 
month. No nntrke^l lishes were* taken in I)t»ccmber. The reci»nl of one haul m i>o«d 
E for Septenil>er is given for comparison with one ma<le in pool \ on thesiimc date. 
It shows that siKHMuiens fnmi \HHil E were smaller than those fi-om pool A, which 
was the deeper pool. 


The iiiiiiii vrthu' of rliis tables i» in sliuwiii^ llio sizi* of the yoiiuK «aliriyn in 
dilTwi'iit months. Tlic jivcmjto siw in AiiKUst wns :i.lil iiiHu's; iif the Aiit;n»t- 
marked Hpeciinons lukcii in Se-jiti-nilKT, ■i.M inchi'rt, n vory «lij:hl ili'crwtMi'. It. will 
1*0 notcMl lat«r thnt thoi**^ was a scarcity of fiHiil in Scirtcnih)-!-, which wuii Id h<to(iiiI. for 
a Bli^flit migration. In <>ct')lH<i' the avLTa^^c sizi' of llic marked siiccimciiH increased 
tn4.20inchcM, l»nl ivmiiiiKHl Ihe same in Nnvcniln-r. In I he one numth, Sf|il<-inlK>rl8 
to <)ctoI)cr IK, thoii^ wa.san ini.' of o,;u inch, l)nt- the tola! increase for the thr«i 
miiiitlis, Angurtt 18 to NovcnilM'f Is, wasbnl O.l'Ip inch, or o. Ill incli per month. This 
small IncrertHc for the t^>tal jHTiixl indieales either tliat tlie growth was very .hIow or 
that the ini^i'ation during that time almnsl coiniieiisati'd for lh<' growtli. During 
SeptemJH'r, pi-olmldy owing to n se-arcity of foiKl, llic migration was a lillie gn-ater, 
and >inn i-esult the fishes wci-e smaller Septemlicr IS than they wcreamiinth l»efon'. 
Thcw'-was a decii-ase of II. (1 inch in Decemln'r, when nearly all ha<l h-fl the jskiI. 

Krmn measurements made on siM-cinien.x Ijiken at (Hcma, we have tlic following 
table for ^e and size of young salmon ivaiaining in fresh water: 

The growth in fi-esh water is, therefore, very slow, and in arliKeinI i>roi)agal ion 
every effort shonld be made to piwent their iviiiaining in the river over summer. 
The growth in salt water is much more rapid. Thi^ saliuou should reach the ocean 



when alx^iit 3 iiK'hrs loii^, and grow to In* 'M inches in iwenty-fonr months, which 
wouhl l)e an increase of abont 1.4 inches i>er month. 

(insfrir pniuisif4 s. — Of 2n:> frcsh-watcr si>ecimens examined in tlie investigation 
of fo(xl of yonng salmon, '-W lia<I parasites in the sttnnach. T\w parasites were of two 
or tlirec kinds, oik* eloniratcd, the othei*s short and grain-like. Tliey have not lKM»n 
stmlied, c-\c4'pt to note the date and siz(» of the li.sli. It is evident that residence in 
fresh water is conducive to the growth of parasites in the stomachs of young s^ilmon. 











jii^M with 








Si /.*.'. 

1.4 \<*'i in«-hi-s . 
'*.\ To '4 ini*h»*4 . 
•i 1 t » 1 inches 
4.1 to 5 inches . . 
5 1 tort;; in«hfs 

Total ... 






a^ff with 









, OctolK*r 


De^'ombor. . 











Total ... 



Dismstil parrs. — Only two diseased young siilmon from the streams have l)een 
met with. One was found dead, covered with fungus, near Sims in lSi)S; the other, 
5 inches long, was taken in th4* trap at HattU* Creek, September I's, ItMX). The upi>er 
lobe of the cauchil fin was wanting, and the remainder, with tlie caudal i>eduncle, was 
coveivd with fungus. 

Mature iiuih pnrrs. — In ()ctol>cr, 1S1»7, sevt^ral mature males, lK*tweiMi 4 and 5 
inches long, were taken at Sisson. In January, 1S'.»S, two males, 5.5 inclies long and 
known to Ix* onlv a vear old, were taken alcove the Hear Valh*v dam near Olema: one 
was mature. In August, 18'.»s, a 4-inch mature nmle was taken at Sissou. Four of 
the 6 young salmon taken at Fall River Mills in August, ISUS, were males, all with 
the genital organs mature. Mature male pari-s were fre<iU4Mitly taken at Uattle Creek 
fisheiy in ()ctolM»r and N4»V4»mber, isus. The sex of a number of parrs, 4 to ♦) inches 
long, from the general collection was4letermined; 15 were mature males, 2 immature 
males, and 11' wer4» f4Mnah\s. Tii(*se mature male parrs can usually Ik* distinguished by 
their more duskv color and bv the slightlv distended alMlomen. Examined under the 
microscope, the milt is a[)parently the same as that from a<lults. A fc»w eggs from a 
female of ordinarv size were fertilized bv milt from a 4.7-incli male. Tlie fertiliziition 
was complete, all of tlie eggs hatched, and the alevins were of normal api)earance. 

No explanation of this early maturing of males can l>e made, and nothing is 
known of their future history. They feed the same as other young salmon and appar- 
entlv are not attracted bv mature f(*males as the adult males are. It mav l)e that 
they return from the ocean as tlie stunteil form known as grilse. It is prolmble 
that several months' residence in fivsh water causes the generative organs to mature 
both in the adults and in tin* young mah»s. 

Temi>eratnre notes. — The following table shows the number of young salmon 
taken at one haul of the seine in water of various temiH»ratures. It indicates but 
little, except that young .Siilimm may reside during the summer in water having a 
temperature of t)4 ^legrees. The 25 taken in Thomas Creek, with a temperature of 
08 degrees, were landlocked in a shallow [>ool. 


Tubh' iif iH'iiu--liinil mill iiiilrr-limiH-nifiin- nvonU. 
ihitp. .vf- I w^ ,Vi' :*y r.:- w- :iy «■■ m- ii.~ itf iw j ii- 


Ill-Pit . 

"ilv.. .'""!'.;''"■. juij- 

D.. Il.t. 

ItodiiiiiK MiLy 

IMi Muy 

I)...- July 

D.. Ann, 

TlHtt!.- t'n-oir Apr 

Di... .IiUy 

Iti-rtlJuff Ajir. 

H>'l<>ir KtilUutf .lulvK 

Miiiitli of Tli..niah I 

fni-k May 

Vlnit. - May 

Jn-lntii ! Miiy 

ItrlnifH May 

HlxVlHta Msy 

B-ufa-in May 

Siui Pnblu Bay Jimo 


Cmtrhisiim ilr-iwii- from uluihj of Niimimr /(.■</'/(/(/«. — It si.h-iiis i>riiloiit' from 
thwi* oliMcrvatioiis tlijil lln- latcf fiy that hat^-Ii in tlii; hi-Hciwulcrs, or nrt' plaii1<>(l 
tlieru jift<T IIk' sitriii;; Iri'slicls liiivc- ]»(ssf<l, ai'i> lialtU; lo ivruaiu till 11i« raiiiN of Ilic 
followiu;; winter. 'VWih ini.'ans it hUiw ^i-owIIi fur at kvisi <> nimiDis, or alioiit ii fifth 
of Ihcir Ki'^wiiirf |MTioii. It iiu'aiis IIr- invfiH'ioiiMiiia(iiriii{r<'f iho laah's, wl)i<-h iimy 
Im ivspoiiMilihr for thr j^ivnl iiiinibi>r of <1\varfs known as ;;rilt(i-: an<l it inoaiis llnil 
I."> iK-r cijiit will Ix-conK- iiifeslod wilh fiastric pantsilt-s. For llit-si- ivastiiis it is 
iinjR'rative that Ihi- fry fiiini our hal<'hfriL-.s Mhoiilil not Ik- rt'U-ascii alxnc Itcililin^ 
after th<' sprin;; fri-shcts, though thiy may iic ivli-ascil in lUe hradwatoiTi carliur 
witUont any <1i-1rinii>nt, anil th«'y ci-rlainly slxaihl mil Im- lii-lil Jiflcr tliis linu- iiu'ivly 
for the i>ur]mM-of fi'iilinj,'. SuiH'rinicmU'iil Slichhy of Sisson Malvhi-iy Mtat4'S, iw 
this ]m]ii'r is ^oint; In p^-Hs, that thett- an- not noarly no nnuiy yoiiii;^ salninn 
n-mtiiniiijr in the SavrHnii<nto Itivi'r nciir SImsoii diirin;; llii' sunnncr sim-t- In- huH 
(|Uit Inihlin;; th^ fry in thi- liat<-ln'ry daring lln- cipring fur f<'(><lin;r as llicn; wcm 
wht-n lie <liil so hold tlivin. Thoif is mi ailvantagi' in hohliiit; fry in tin- halclnfrius 
for ft'wUuy. 


fri'iifrril uliiihj iiffimil 'mfri-xh ii-iih r. — The yoiiny siilnmn fi-cd principally npon 
floalinjr or driftinf^ insocts. either imnialui-e or adults. When foedinji ihey often 
taki- a Illation below a stii-k or rock and catch their food as it (loals down on eiilu'r 
side. TheyeaK»'rly eateli snuill in^iH-tsand larva' if thrown intolhv water. Fry 1.5 
inches lonfj; have l)een olwrved to rise- io a small lly Ihal ali^rhted on IlK'natin-. They 
hardly ever eat eneased eaddis larva', n1thou;;lt that is I he main fcHid of the troiil. 

The folhtwing is a tiilmlar stalenieiil of the stomach contents of lii.'.'i yoiinfr 
salmon, Ix-inR Imsed on an examination of alsinl five specinu-ns from eaeh liH-ality 
eauh month in the year in which any were taken. Thi" rcconl fori-aeh lislicxtiniincd 
oon»islK uf t)ic Stat ion, date, sixe of fish, and niimlK-r of specimens of eaeh kind of 
fiKMl or other material found in the stomach. Four forms of inseeis an' n-co^iui/.ed 
in the talile, viz, larva'; pujHe, includiu}! nymjihs: flyiu)^ insects; and "lenvstrial 
iiusovls," iueludinj^ adult wingless iniK'ets and spiders. 



Table of Httwiarh ronienta of yonnij sttitiion, Sdcntmento Daain. 

+ indicates prosoni.t* of ctTtaiii olijivts. tlu* miinlMT t»f whi<'h was u«»t iU»tormin«.'<l. The totals iu last two 
oolxunns indif-at*? numlx-r of flsht's in whose stoiniK.'hs iiarasites nr in<li>f(*stiljl«' miittTial whs fnuiul. 




•• • 


X \ 

ti 7 

X«». of SlM.H'i- 


3 , 

.» ^ 12. 

Sulla way Cixvk, May 15, IH!«s 


















.« fr- 

C" W w 


r-' 5^ £ 5: 

? -! 

1 I 



Total Hi rt 

A%'erage •^- 'i 

Sulliiway Crv!««k. Auk- 525. 1^»^ 

7 HI 

f< :il 

«.♦ H. 8 

III :15 






1 1 


a» ... 

2 1 .. 
1 1 ' 

Total !«.«» 

A%-era[^e 3.*J 

Siiw^ui. Aiiir. r.». l«w 

11 :J. 1 

12 ' 3. 1 

13 3. 3 

14 3. 5 

15 3. 7 

Total 1«».7 

A veratf i» 3. 2 

Dunsimiir. July .S l^(«^. 



17 3.0 

IS :i.M 

\\\ 3.H 

Total 1-M 

A verajfe 3. <» 

Siin>. Mav 17, l>!fri 

ai , i.r. 

21 • 2.«i 

:i* 3. 7 

•/3 i '.\. •> 

Total ■ 11.7 

A vtTUir*" - ' - *•• 

Siiu-.Julv '.». W.«s 



H. 5 
H. '.♦ 














4<) ' 

3:^ h\ .... 1 

7 3 





4 : 

3 1 



* 1 
.1 • 


2 ■ 1 ■ 






1 3 : 




is . 
1 2 

1 1 







1 : 

T..tal l»i.U 

Avcrajfo 3.4 


Sim.*.. Autf. 17. l«fr<. 

21^ . 3.H 

.HI i 3.4 

31 3. 5 

:« 3.6 

'X\ ' H.h 

Total 17.6 

Averajfv 3.5 





■ Sini.**. Sept IS. isi^ 


3 7 

4 5 


Total ' 12.0 

Averatre 4.0 I 

Sinis. i^X. IS, ISUe 

;I7 3. 1 

:^ 3. h 

;fi) 4 1 

40 4. I 

41 4 tt 

Total I» 5 

Averago 3.11 



5 .. 
I ■ 

16 1 










I ^ 











I ■ 

I ■ 

1 1 

■ I 



t : 

; I 


Table of stnnnwh c(mtentH of ijonntj wthm^ii, Samviwnto liasiii — CV)iitinne<l. 











« 5 







^ *^m 



U) f 





.3 t 




•Z u 





c 5 








9 * 

i i 

* i 

- d 

SllUH, Nov. LH, iKflrt 

Sim8, !)««<•. IH, 1««K 


Haz<»l <;roek, July «, IWW 


Haz«^l Creek, Sopt. IH, 1«W 

Averau:** . 

Hazel ('riM»k, ()<-t. IH. 1«« 


No.of SlKM'i- 

iiinnM wit li- 






Hazt'l Croi'k, Nov. is, 1898 


Hazel CYjH'k, Dw. IH, 18HK . 


t'all Rivor at Dana, Auijr. as. IHIW ... H\ 
Fall RiviT at mouth. Auk- 5fl', 1H9H 


McCnoud Rivur at B«iitl, S««pt. 16, 


RtHldiuK, May I, l«i8 


nati;kai. history of the QriuwAT salmon. 


Tahlv ttf tttoiiuwh nnitvuta of t/tmrnj salntou, Sdctftinenti* Hasin — Coiitiiined. 














- • 





















« 1 








*■ ■ 

No. of S J UN" i- 

ineriK with— 

... - -f" 

Re<1(iin»(. July 11. 1^4#* 







" ~ 

™ — 

1 ' 

....; 1 .... 

Total ■ 

A vomcfc* . . ! 








■■-" ■ 


— _ — 


KalLs Ferry. Jau. 21, 1«W 

T« »tal 




2.5 1 

' 1.5 ■ 
1.5 1 
1.7 ■ 

1.7 ! 



1 A vcrag'*? 

....' • 


W 1 




— — 

Balls F.-rry. Fi-b. \M. IMW 

■ ..1 

....'. .. ■ ...1 





7.t» 1 


<. 1 











A vera^i* 

- - 


= ■, 



+ ' 

■ Balls Fttrry. Mar. Xr*, 1h!W 























I ...I.... 

















BattleC'rv«'katI^inK>*.Sei»t.l4. 1.s'.is. 










: •.-_ - 


Balls Ffrry, July VA. \t<f> 


- • 





















Battl.* CnN'k. Apr. :«». IXw 















12. It 






. m m m 


Battle CYifk. Nov. 4 to i:*. 1«»* .... 

f i:*i 




; 1 

1 + 










1".... 4- 

• *« • , 











3 -. . 

1 Average 






Table of stitmach cnntvtitH of ijoumj tuihmm, Sacramento Basht — ContiimtMl. 











es $ 







w ••^ 





/ 11(1 


















Bat tlo Crook, < Vt. « to «», Irtw . ... 


















V i:«> 


Total - 








RtHl Bluff, May 21 . !«»* 








10. H 






Chiro. May2:{, 1H|I8. 



Jiuinto, May 21, iNiiK... 

Tf >tal 


Butt»(Mty, May25. IWIH 


(?oluHa. May 27, ItW .. 



« Yulk not yet all abttorlKKl. 














m I 






r l(t7 
\ 1«H 

I 111 
\ 17( 














•» 1 























2 ... 

3 ... 
1 ... 

4 ...-i 
15 13 

N»». of HJNfi- 


X I 

e 1^ 


■ X 

* s 



S *x •e'^ 

.- I C 1 5 

5^J £ 16^ 





1 ....' 


: f 
' '+ I. 







8 : 
20 I 


13 ...J 

1 ;....'. 

i I 

5 I 1 I 1 
I I 

1 1.. . . 


i" .... .'.'. 


1 I I 


4 '. 


'.'.':• 1 ' 

... + 








10 j 3 
4 ' 20 
8 2 

28 I 25 
. 8 

\i)' -.7... 

:w . 

13 . 

f I. 

''Salmon eprf;, wiMti^ fn>w HjiawiiinK platform. 


TViWi' iif xliiiimcli ciintenl* irf i/iiiiiig xahnoti. fUtcramenIo liimin — Continn^, 




i i 

5. t 






(U .... 

■i I 




- — 





f i 

T.,»l_ . -, 

_' ; 





;:::;: ^ 

:::: : :;.: 


1 B 





- ^ 



T.rtnl- - - 


^'^ ^ 


I ::::: 

........ 1 






1 iw :« 

ail sj 




■■--1 ^ 






Amnei ...... . 

Bi..ViHl-,Mi..vll8i.rt:«i. I«« 

;;;:;: \ 

rr ^^ 



— ■^ 





^ -;^ 



^ ^ 


+ 1:.: 

Tw* 1,™ 



— — 


llrnrral tnmmnry -f frrnh-irulrr 

;»7« K an 



Si 12 





Tliis slndy shows titat yimiiy siiliiiiui in fresli water fttwl exclusively on insects, 
ami tliat iiuiiialure atiuiitiu inseets Tonu hyfitrthe larjjer portion of their food. The 
>:euei'Hl stimmaryof thelal>li> shows that ii[)pm\imatoly lialf of the food of thcsiH>ci- 
ini-ns studied c-on.HisUHl or piipm (or tiyintihs, whieli were not distingiiislicd fi-uiii 
t)ii[ue), one-third of Iiirvie, niu\ oiie-sixtli of adiill winfred iiiseotB. 

Tliero was an increase of flying iiiseels in the ffKul of speeinions tnken in Sull.i 
way Creek in August, and anineream'in anion ntof food in speei mens fi-oniSinisdnrin: 
.luly and Aiignst. It was during September, when apjKirently lliere was » scarcity 
of footi, that the larger young salmon disappeai-efl from Sims, Ther«^i wasan inen-ase 
in flying insects in food of siieeimeiis fi-<)m Hazel Creek in SeplemlK'r: a scairity of 
foot] and a noticeable hick of larvie in sp<><-imenH from Itnltle C'n-ek in Oetolx>r and 
November, and a smaller amount of foiKl in specimens from the lower stations. (See 
summary for May, i>age 117.) 

Two spec im«ns from above Hear Valley Dam, near Olema, taken January IS, 
ISUS, had stomachs gorge<l with larvse and pnpa-, one having al>out 5iiof Ihe fnnner 
and 25 of the latter, but no indications of adult insects. 

Tl live specimens, Nos. 117, IIS, and i;i7, of foinl table, wei-e taken in <)ctol)eranil 
NovemlH'r before the yolk was yet altstn'lM'd. One had nothing in its slomaeh; 
another had some food, but it was unrecognizable; Ihe thinl had ealen one larva and 
Iwo adnll inseeto, Iwsides senile other foiHl that was unrecognizable. This indicates 
that they liegin fewliiig even Ix-foi-e the yolk is all al>sorl)e<i. 

The food datji, if arninged aecoi-ding to size of (isli, wouhl give the following 
average amounts iH'r tisli. This table shows that pnitie and nymphs an' the favorite 
foixl for all si/.es. Those fmiii lA to - inches in leiiglh feed very little ujion adnll 
iusoc^t^; the largest size feed very Utile tiixin iaivie. 

The following table brings together a slalcnu-nt of Ihe average amount of fowl 
found in the stomachs of the young aalnioii from various stations for the month of 
May, Ihe only month in which we seeui-ed young salmon fvom many of Hie h)wer 
stations. The table indi<'a1es that theiiniiorbint foiK) of the young salmon through- 
out the basin in May was lan'fc and pupte, of which there was an average of 4.4 of the 
former and U 5 ot the latter |«*i fish. The fish examined averaged 2.5 inches. It 
also show.s there was a slightly smaller amount of fotMl in s|>eeinieiis from the lower 
portion of the river. They weie not starving, however, and there is no evidence 
that the passage down the river is detrimental on a<.-connt of Ihe lack of finKl, 
" in<lieates an average of less than one. The nuinl)ers in the eolnmns hcadc<l 
"No food' and ' Paiasites'* indicat<! the numlM'i of lishes e.xamincd ihat had 
empty stomachs or parasilcs, as tlie ease may In.'. The totals aiv taken Inmi Iho 
eoinplele tabh^ ol lotHi, hut only for the month of May, and ai-e not the sums of the 
averages given in this table 



Table shoiriug the aventg*' amount of food in stomachs of young salmon from the 

i-arious stations in May. 


No. of 

I s]>ec'i- 















"^fi^" Cru^ta (Tftstrt^^ 





Snilaway Creek 


•* •> 

2 2 

9 1 














ft 1 

i<ii t*i» (a I 



Battle Cre»'k fishery 
Red Bluff 

1 '1 1 






Chi(X) Bridjre 














("t ..... 

Rio VLsta 






2. 5 


»i. 5 

«•> 2 1 

.1 i<i 1 ,ti , 



•» Indi»'at«*s an av»*raK»* «»f l<***» than on** 

F<n>i1 In Ifraf'kish trater. — Rehitivoly few si>ecimeiis of y4)un«^ salmon have been 
obtained from brackish water, and tlie following table irives a list of the foo<l found 
in nearly all thai were canglit: 


Benicia, Feb. 21 t«. Mar. 3. 1>^». 

N.'. Size. 

d« 1.8 

flC 1.9 

aK 1.9 

2lW 1.9 

2HI 1.9 

211 2.2 

Total 14 3 

Average* 2.0 

Beniria. Mhv i:i-J». l.'*.*^ 

Amphi Ome pj^j^ Adult ^^^^^ F>ra- 



San Pablo Bav.Mav 17 21.1>«»<. 

Total 211 2 

A veraife 

SumiHary/ttr Jf) brackish-init* r x^KTiMini*. 



The chief f«HMl of the few brackish and salt water specimens studied were adult 
inse<*ts. Only 5 of the 20 siHH'imens had fe<I to much extent on copeixxls, and only 
1 had eaten a fish; the six^cies of the fish could not be determineil, though it was 
evidently a smelt {Osinerus). No aquatic insects were found, such as were found 
in specimens fnmi the lower river, which indicates that the fish had lK*en in brackish 
water at least long enough to digi*st all fresh-water food. 



Ciassificaliim of lh€ insect- foot} of young s(i}moii. — The foUowing is a classified 
list of the insects faiind in the stomachs of young salmon from the Sa<-nimento 
River, collected during this investigation, as reported by Miss IJerlha Clinpman. 
It is evident that in many c»r<>s the fish confined ttieniselves largely to fowl (^oUert^d 
at the surface of the stntam, as is the case with fish taken at Rio Vista in Ma,\', or 
those taken at Fall River Mills in August; otiiors sought the iinniatnre forms living 
under water, as can be seen from the majority of eases in the list. Jlnl in no ease 
can this distinction in feeding habil.s be definitely miide. They seem i<> have fwl 
indifferently on water and surface forms. These surface fonns aii' almost invari- 
ably insects living almut streams, and which might thereftn-e easily have fallen into 
the water from overhanging plants. Mueli of tlio slonuK'h contents had Iwen so far 
digested that it was not possible lo identify the insects. Other ins«'cls have l>een 
partially determined l>y single wings or particles of the Ixxly; but it ncenietl net so 
important to cany the classification to siweies as to detennino the tyis-s of insw'ts 
forming the fooil of young fish. The resulls of the study are given in the following 

(HaMKifiFutioH iif iniui-l fmxl/tir ynuinj nahiiiiu. 

Balls Fonr.. 
BallH F.Try . . 




Bulto City . . 

Plniils rhrw^khllil. 
Point nli'hiutind 

Buel'Crvi'k . . 

tfaxIdloK - 





^ ' 



? . 1 




Claxtijieation o/ iuHect fcnxl for young salmon — Continued. 

I s 

S k 




Walnnt Grove January . :*• 

Balls Ferry February *2 

Walnut On)vo do .... ♦^ 

Benicia do 3 

Ball8 Ferry Man»h .... *Si 

Walnut do. ..♦ai 

Benicia «l>ra<'ki>h ■ do 4 

Battle CYeek April 5 

Sacrament4> do *\s 

Olema do . . ♦tis 

Si!«K>n May 5 

Sims do 4 

Bedding ... « 

Red Bluff do ... 5 

Tehama do *> 

Jacinto do 2 

Rio Vifita do ;> 

Benicia ( brackish ) ... do 5 

Pinole ( brackish > do 7 

Dunamnir do 

Hazel Creek do 

2 I 


1 I t 






1 1 






Redding do 

Battle Creek do 









."i 2 

Siseon Aujrust... 

Dana do 

Fall River MilL* do... 

Sims do — 

Sullaway Cre^^'k. . September ♦IS 3 

Sims do ♦y ; .. •: 

Baird do 5 ' 

BattleCreek do... .MS ;/3 li .. ♦ 

Sims O-tober.. 5 

Hazel Creek do '> 

BattleCreek do... a> 1 

BattleCpcH»k Noveml>er 10 i... 12 

Sims December .*» '... 1 


.1 1 








■ Me^ns pn»sent, but numb^-r n«»t kn "wn. 
/ Means larva. 

;* Means pui»a. 
(f Means adult. 

♦ Mean.** mori' than li». 
X Means more than 5. 

The f<)ll<)win«r list of eominon iiiimos will 1k» of assistan<*«» to those who are not 
familial* with the seientifie terms used in tlie alxive table: 


Libellnlidte Dragon-flies. 

Ephemerida . May-flies. 

Plecoptera ... . ..Stone-flies. 


Acridi<liP Grafv^hoppers. 

Hemiptera Bngs. lice, aphids. etc. 

Corisida' Water-boatmen. 

Notone:-tid-e . . Back-swimmers. 

Capsidie Leaf-bngs. 

Jassidte . . . Leaf-ho])per5». 

Membra<.*idie . . . Tree-hoppers. 

PsyUidie Jnni])iug plant-lice. 

Apbididie . Plant-lice or aphids. 

Trichoptera Caddis flies or caddice 


Lepidoptera Moths, skippers, butterflies. 

Coleoptera Beetles. 

Staphylinida* . . . Rove l^eetles. 

The following is a list of insects having the early .^^tages pass^Ml in the water, 

select e<l from the set worke<l over: 

Ephemerida. Trichoptera. Chironomida*. 

Plecoptera. Tipnlidap. Simnliidap. 

Corisidee (entire life in water). Blepharocerida*. Leptidap. 

Kotonectidje (entire life in water ) . Cnlicidse. Coleopteroos larvae. 


Ti])nlid5P Crane-flies. 

Blepharoceridie Net-wingeil midges. 

Cnlicidie Mosqnitoes. 

Chironomida* Midges. 

Ceratojx>gon . . . Pnnkies. 

Simnliidie Black flies. 

Leptidae Snipe flies. 

Empididae Dance flies. 

Phorida? Hompbacked flies, 

Mnscidap Mnscids, 

Dexiime Nimble-flies. 

Anthom\iina? . Anthom\iids. 

Mu^-inap Common flies. 

Hymeno]»tera Ants, bees, wasps, etc. 

Ichnenmonida? . Ichneumon flies. 

Brax.'onida^ Braconids. 

Formicid* Ants. 



Of the ocean life of the salmon very little is known, although it comprises two- 
thirds of their (existence. They have l)een taken in Tomales Bay when four months 
old, about 20 miles from the mouth of the sti'eam in which they had l)een plantt^, 
and a few of al)out the same age have l)een taken in San Pablo Bay. These were 
on their way to the ocean and were alreatly in nc^arly pure sea wat^^r. SiKH»imens 8 
to 15 inches long are sometimes taken by anglei's in San Francisco Bay. In the 
estuary an<l along the bear'h at Karluk, Kadiak Island, Alaska, the writer has seen 
schools of several hundred S-inch red salmon that had come inshore with the adult«. 
These wen^ f(»cding and were not dwarfs on their way to spawning-grounds. 

Salmon an* sometimes taken iu the o(*ean near San Francis<M> in ])aranzella 
nets and are also captured in large numbers with the troll in Monterey Bay, wheiv 
they api>ear in February and are found until the mi<ldle of August. 

Something of the ocean life of the salmon might 1m» learne<l by making a study 
of the food found in the stomachs of individuals taken when they first enter Mon- 
terey Bay. At such tinu\s a portion of the fo(Ml taken on th(»ir offshore f(»edi ng 
grounds should yet Ik» in identifiable condition. Tlie presence of deep-water fishes 
or crustaceans would indicate a deep-water life for the salmon. 

During its life in the sea the salmon is, of coui-se, not entin»ly free from 
enemies, and something might be learned of them by studying the semi's left by the 
injuries they inflict where they fail to kill. Injuries and deformities received befoi-e 
entering fresh water were of frequent occurrence* among spawning fislu^s at Battle 
Creek in 1S97, but no particular attention was given them. Tlie males oft<^n had 
the snout twisted or split, or even cut off. Very often there were one or more scars 
ext-ending obliquely backward and downward on the side al>ove th(» anal fm. Some- 
times two or three were parallel, as if they were scrat<.»hes made by teeth of some 
other fish while the salmon was smaller. These scales wcmti more oft<?n present on 
the femal(»s. The dwarf females were always injured in some way. Very few 
injure<l fishes were observed in 1808. Two had lost th(» ventral fins, one had lost 
the lower two-tliirds of the opi^rde, two had defonmMl backlnines. Only one fish, a 
female, had the oblique and parallel scars mentioned; th(»y were slightly curved and 
in two s(»ries of s<»ven each. This subject is worthy of further consideration. 

A very chara(*teristic scar, and one that always attracts att^Mition, is that left by 
the sucking disk of tlie lamprey. The lamprey has no lower jaw, its mouth being 
circular and of the size of the end of its head. The gullet ends in the middle of the 
mouth and is bounded on the upper and lower sides by hooked teeth. There art^ 
other smaller liooked te<ith above and below these on tlie disk, an<l on each sid(»of tlie 
disk there ait^ about four short cross rows of teeth, and the whole circumf(»nMice of 
the disk is beset with small teeth. When the lamprey att acinus itself to anotluM- fish 
the outer row of small teeth leaves a scar somewhat resembling the milling on a (»oin, 
which has led imaginative i)ei'sons to see in the whole scar a resemblance to a brand 
made by a heMed coin. The illustration on plsiie 1.'3 is ma<le from tlic^ photograph 
of a lamprey scar on the opercle of a blueback salmon found at Karluk, Alaska. 

Ill 1800 5,000 marked fry were released in a tributary of the Columbia River; 
about 400 of these were taken in 1808, and a few more each in ISOO and lOoo. This 
indicates that most salmon remain in the ocean two years, though a few remain thre(» 
or four years, as will be seen from the following chapter. 




Do salmon rfhirn fo fhfir mtfir* sfrtfims/ — There is a widespi'ead lH4ief that 
when a salmon n»turns to fr«»sh wat<»r to breed it seeks the stn»ani in whieh it was 
hatched, thon;Lrh th<M'e is v<*rv little evi(k*nce that such is true. Various lishermen 
claim tliat tln\v <'an distiuiruish the salmon of parti<*ular streams by tlieir »reneral 
ai)i)earan<'e, which is in<*nMlibl<\ Th<» employers of tlie Alaska Paek<M*s' Association 
stat«* that the re<l salmr»n taken at T^anuk an* always smaller than thosr taken at 
Karluk, lM)th i>laccs on the north coast of Kadiak Islan<l, Alaska; that 13 of the 
former ar<» nniuin'd t<» make a case of <*anniMl salmon, wliilc only 11 of the latter are 
nec<»ssiiry. This M^ems to indicate that the salmon of the two localities are distinct, 
])Ut tlic lar<rer salmon may «;o to Karluk, m»t because they have be<Mi hatclied in 
Karluk Lake, but IwM'ause thev are larirer. 

In 1SU7 s;».">,(KM) ((uinnat salmon fry were release<l in Paper-mill Cnn^k and its 
tributaries inl<» Tomah^s l>ay, California, and 2,<mm»,(Mm> alevins were 
ndeasiMl in the same stn'ams in ls!»s. (See ''()bservali<»ns on alevins aKificially 
ivan*d.") In T'oo a few salmon wtM-eseen in Paper-mill Civek, an<l in lind they were 
abundant. In one liaul of the seine in the tidi»-water portion <»f Paper-mill Oeek, 
coverinjj: a section about l.">o feet lonjr, " qiiinnat salmon were taken November 1(5, 
llMd. It is wtdl known that ([uinnat salmon did not breed in Pai)er-mill Cn^^k or 
its tributaries previous to ls!»7, f<»r wliich reason these streams w(»re sele<'ted for the 
expi»riment. Mr. Th<»mas Irwin reixu'ts that h<» saw two larjre salmon in Pai)er-mill 
Crei'k about isnn, Imt witli these excentions he never saw anv fishes in tlie sti'eam 
that mi^ht be taken for ([uinnats until 11hm». lie lives on the banks of the erec»k 
and knows the stream thoroujrhly. His statement ajj:rees with that of other ixM*sons. 

I^H)er-mill Creek is not suitable for quinnat salmon, bein»r entirely too small, 
but it is frequeiitiMl by do^ salmon and steelheads. 

Hut there is no conclusive evidence that the fishes whieh were fouml in Paper- 
mill Creek in I'Joo an<l I'.'Ol were tlie sjimc individuals ndease<l there thi^ee or four 
years previously. Tlu\v may have been merely stray lishes, and tlieir In'injr found 
there at tliat time onlv a c<»inci<len<'e: or their cominir into Tomales l>av mav have 
lx*en causi'd by then* beinjr «ni extra larj^e number of salmon in the o(*ean, which 
mij^ht very well be, owing to the larj^e output <»f younjr fi'om the halcheri<'s; or thoM» 
foun<l in Pai)er-mill Creek in llMK»and lU'd may have been some* of those ndeased 
then\ in whi<*h <*as(» it is very probable that they ha<l nevc^* reaidied the ocean at all, 
but remained in Tomales Bay. Paper-mill Cn*ek would th<*n 1m» their only stream. 

It is incredibh* that the salmon remem]M*r their native stream during their two 
or three vears of <»cean life and that thev consciouslv seek it when tht»v <lesire to 

• • • • 

r«»turn tt) fresh water. Pi-obablv most of them do return to the stn»am from which 
they entered the ocej^n, not because it istheir native stream, but iM'cause tlu\v do not 
<r«'t far awav from its mouth, an<l when readv to return to fresli water it is the first to 
attract them. 

TJtr inn runs nf siihnnn. — .Vdult salmon mav b<» found in tlu* Sa<'ramento Kiver 
at almost any time of the year. There are, however, two more or less distinct runs, 
till* fii-st of whi<*h passes up the rivei- dni'inj; Api'il, May, and .lune, an<l the latter 
during .Vuirust, S<*ptemlM*r, and OctolxM*. The formiM* is kn(»wn as the sprin*; run, 
the latt<»r as the fall run. 

The salnK)n of the sprin«r run ascen<l the river to the headwater, such as the 
Upper Sacramento, McCloud River, an<l Hat Creek, and some of the earlier ones even 


pass Pit River Falls and aseoiul Fall River to its source. They are not found in Pit 
River above the mouth of Fall River. By the time they reach this portion of the 
stream, the Tapper Pit River is very low and the water impure, and the salmon all 
turn into Fall River. The salmon of this, the spring run, spawn mainly in August. 

The fall salmon do not ascend the river as far as the spring run, but turn into 
the lower tributaries or spawn in the main river. They rejich their spawning- 
grounds during the latter half of October, November, and th(^ first half of December, 
and spawn soon after. The main river is very low at that time of the year, and the 
portion between Tehama and Redding is an i in ^wrtant spawning-ground. (See chart 
of spawning-grounds, plate 17.) 

As a matter of fact* there is no definite distinction l)etween the spring and fall 
runs; that is, there is no time during the summer when there an> no salmon running. 
First tlien* are a few very early salmon that begin running up the river in Febniaiy, 
and the number inci*eases until May when it dec refuses till July; then it increases 
till the first of SepU^mber when it. again deci*eases, there Iwing a very few each month 
until the next spring run. 

The spawning seasons merge in the same way. The earliest salmon go farthest 
upstream, and as the sejison advances tlu^y stop at lower points. The localities and 
dates of the spawning of the earlier salmon have not been det^»rmined, except that 
Superintendent Laml)son, of Baird, reports liaving se(»n a pair of spawning salmon in 
the McCloiul at the hatchery on the 20th of April, 11K)2, which is the earliest record 
known. By the 1st of OctolxM* spawning fishes are found as far downstream as 
Redding, and as far as Tehama by the first of November. 

Details of mujration, — Wluui the salmon enter San Francisco Bay tliey come in 
against the ebb tide, stem the <Hirrent till the tide changes, and then run out against 
the flood tide, losing much of the distance gained during the ebb. How it is that 
they do not los<^ altogether as much as they gain will lx> understood fi-om the following 
explanation: The tide runs up the bay and river as a broad, low wave, on the upper 
side of which is floo<l tide and on the lower side ebb t ide. When thecrestof a wave — 
that is, slack high wat(»r — is at Isleton, the trough, or slack low wat^er, is about at the 
Golden Gate. This wave is about three hours reaching Benicia and four in reaching 
Collinsville. The farther up the bay and river it rea(»he8 the smaller the wave 
becomes, the shorter the flood, and (as the flood and ebb combined must Cfjual 12 
hours) the long(»r the ebb. 

The following diagram will illustrate the movements of a salmon in passing 
through th(* bays: a, ?>, and c represent the ti<le wave at succ(»s8ive points as it 
passes up the bay, < ^» indicates ebb tide, and w^ > flood tide. Suppose that a 

7ot¥4rd Ht€ au9n. 


salmon ent<M's the Golden Gate, GG, at the beginning of ebb tide, which would l>e 
the most favorable time. His position on the wave will be at .s. If he is able to 
travel up tlu^ bay a,s fast as the wave lu^ will ke<»p his position near the civst, that is 
at s. But he can hardly do that, especially as the current would be very slight. 


and in the l>roa<l bay hai-dly stron<^ enoufi^h for his jjuidance. Suppose that by the 
time he n*aehes Henieia, /?, ho has fallen l)ehiml the wave until he has the position 
at s. It is th(»n slack low wattM*, an<l he ean make no hea<lway. ^(nm tlie next 
wave reaches him an<l he is in flooil ti<le. He will tlieivfore swim back against the 
current. As the wave is ^oin<; up thr bay and he is j?oing down, he soon jrets past 
the cn^st and timls himself in the ebb tide at s". He then turns an<l stems the ebb 
tide, an<l as the wave is ^oinjj: in th<» same <lire<*tion h«» is, he <r(>es much l)eyond 
Henieia, />, lM*fore he again falls back to shxek low water at .s'", and gets into the 
flood of the next tide wave. 

Then» is noway of tracing the passage of the s^ihrnm through the bays, but from 
records made at Vallejo, Henieia, and CoUinsville it seems to require al)out a week 
to reach the mouth of tin* river after thev enter the (rolden Gate. 

Plat<» 14 indicates th<» catch of tish at various places from Vallejo to Sacramento 
for a certain period, and is intended to show the passage of two schools l>etween 
the two places. Each division of the diagram in<licates the relative amount of 
salmon taken at the ten plai'cs during one day, the unit being the average daily 
catch at the given phice. The vertical spaces indicate tenths of the average daily 
catch. By a careful study of the diagram the following points will be notiMl: 

On Monday, April 25, there were few fish taken anywhere, the catch l>eing less 
than the average at all [)oints. This is the more marked because the Monday 
catch is on an average 25 ix*r cent gr(»ater than that of other days, on account of 
there lieing no fishing on Sunday. On Tues^lay theiv was a big catch at Vallejo 
(3.0 times the average), and a slight inci*ease at Benicia (1.1), Dutton (1.2), Black 
Diannmd (1.2), and CoUinsville (0.9). There was little or no increase* at other i>oints. 

On Wedne.s<lay, the second day of the run, the catch at Vallejo had fallen off, 
and by Thui's<lay the run had entirely passed that point. The points on Suisun 
Bay and along the river as far as Islelon were gained on the second and third days, 
and the run reach#»d Courtland on Fri<lay, the fourth day. Then* was no fishing at 
some of the upjKM' stations on Satunlay, that is, Friday night, th<» law prohibiting 
fishing fnnn sunrise Saturdav to sunset Sundav, and the n*conl for tlie fifth dav is 
incomplete. I'his run was two <lays in passing Vallejo, and four days in going from 
Vallejo to Courtland. 

On Friday, April 20, another run began pas.sing Vallejo, the catch IxMUg over 
three times the average, an<l the next day it had increase<l to over six times the 
average. On Monday the Vallejo catch decreased to 1.1». on Tuesday to 1.7, and on 
Wednesday to n.4, the run l>eing five days in pa.ssing that place. This new run 
was not noticed at the other i)oints on Friday, but on Satui-day, the second day, it 
had reached all jKjints up to CoUinsville at the mouth of the river. By Monday, the 
fourth day of th<» run, it had reacluKl all points from which we have reconls, the 
gn»atest increase Inqng at the stations farther u[) the river. During the ivmainder 
of this week the catch <'ontinue<l to fall off at the lower staticms, but continued very 
large at Sa<*ramento. By \V<Mlnes<lay, the sixth clay, it had passed liio Vista, and 
Walnut (irov«' bv the seventh. On Fri<lav th<M'e was still a big catch at Sacramento 
(5.1») and at Courtland (o.S). Th<* leconl is imi)erfect for Satunlay as usual, but 
apparently the run ha<l passe<l all stations. To sumnuirize: This run was five days 
in passing Vallejo. The foremost wen* four<lays going from Vallejo to Sacramento, 
and the run was five days passing Sacramento. 

124 BiTLLprriN of the united rtatks fish commission. 

The spring run passes iii)sti'eain ijuite rapidly, reaching their spa wninp:-jcround8 
on the MeCloiid River in about six weeks aft(»r entering the river at Collinsville. 

The fall run moves more slowlv. Thev are about two months n.»aching their 
spa^vning-grounds, which are not so far upstream. The Hood and ebb tid<'s are more 
nearly equal, owing to the smaller amount of water coming from the rivers, making 
the i)assage of the salmon through tin* bay a litth* l<mg<M\ The nets of the iish- 
er men also otter a gn»ater obstruction during th(^ low water and in this way hold' 
the salmon back. In 11)00 salmon were taken in abundance in Suisun Uav and 
in the river as far up as Rio Vista by the middh* of August, but wen^ not taken 
at Sacramento until after the first of September. The low water doubtless nnwle the 
movement slow, and the taking of from 2,<M)n to lOjMM) daily out of a slow run would 
a(*count for their nonappearance at Sacram(»nto. 

Upon n^iching tlie shoals in the mi<ldl<» portion of tho river they <*east» their 
migration, having already found goo<l spawning-grounds. In ls!>s, ISiM.), an<l IIMM) 
the water wiis normally low and a large projjortion of the salmon found spawning- 
places in the main rivcM*. Tin* <»arly high water and fr(»<|uent. fall rains in 1S07 s<Mit 
them into th(» tributari<»s. 

Tlie latter part of September, 1Ih)1, l">OsalnH»n we?*e weighe<l an<l branded witli 
serial numbers and n^leas(»d in the river n<»ar Rio Vista. Thr<M» of these were taken 
at the hatchori(»s the latter part of November, just at the <*lose of the season. The 
following is a parti(*ular account of tlies(» three sp(»(*imens: 

No. 8, a fenml<^ was branded September L*(), when it weighed ].*5,0:JO grams. It 
was taken again at Mill Creek fisln^ry Novemlu'r 2o, wh(»n it w<'igh<Hl 10, 180 grams, 
having been <>4 days (m the roa<l and having lost l*«> ])er cent of its weight. 

No. in, also a female, was branded September l*-!, wlu^n it weigh(»d S,470 grams. 
It was retaken at Mill Creek November 20, when it weighed 7,H)0 grams, its time in 
passing up the river being 50 days and its loss in weight being IT) iM»r (MMit. This 
siRHMmen was returned to the creek after lM*ing w<»ighed Nov<»mlM»r 2o. It was 
found dcjwl on the ra<^ks 8 <lays later, when it had spawn<»d all but 20 of its ova. 
Its weight liad decrease<l 1,800 grams. 

No. 4»5, a male, was branded S<»[)tember 2o, wlien it weighed 10,OS(» grams. It 
was taken at Hattle (-reek November 2;"), wh<»n it weighed 0,27;") grams, nniking its 
time from Rio Vista <»<» days an<l its loss in w(»ight 2-5 per cent. 

This import^mt exiHM'iment proves that tlie fall salmon travcd very slowly, at a 
rate of 4 or T) miles a day, and recjuire about two months to n»aeh tin* spawning- 
grounds from the mouth of the river. 

The salmon of the spring run arrive' at their spawning-grounds fnnn two to six 
weeks or even long(^r betore they are ready to spawn. This time they spend lying 
quietly in the pools. The fall salmon are mon» nearly rip(» wIhmi they reach their 
spawning-grounds. Indeed, it. is probable that many of tlnMU cease to aseeml the 
streams only wIkmi th<\v ar<^ ready to spawn. 

Downstream morcnnnl, — Tnder onlimiry conditions then* is probably little or 
no downstream movement, yet wIkmi the salmon me<^t with sueh obstructions as the 
racks at the fish-culture stations, there is a ten<h»ncv to go back downstrc^am. At 
Battle Creek fishery more salm(;n are tnk<»n at the lower (Mid of the pool than at 
the upi)er, indicating that they go as far <lownstreaiii as possibI<» under tlu» ein'um- 
st-ances. The large numlKM* of fishes in good (condition that get caught on tin* rack 

April 25. 


indicates :in attempt to go downstream. They may fre<iuently I>e seen eoming 
downstream toward the ra<'k, though I have never seen any try to get tliroiigh it. 
When they get close enough to the ra<'k to feel tin* fon^e of the swift current, they 
alwavs trv to turn ha<'k. Eventuallv thev become so weak that thevare unabk* to 
keep from being carrie<l onto the rack, where many of them jX'rish. 

Tliere are also a few (islies that <lrop downstream as they In^come exhausted 
from long resilience in fresh water; rarely from spawning. Sucli were found ahnost 
dciily on the upiM*r ra<-k at IJattle C'r<M*k fishery in Tjoo, V(»ry few of tliem had 
spawned, though th(\v won- almost <-ompletely <*xliaust<Ml and hardly ever lived over 
a dav after cominir near the raek. Such speeimeiis usuallv lie in tlie less swift 
water some in or 15 yar<ls above th«' ra<*k, wIkmv littlo elfort is rtMiuired to maintain 
a position. As they Imm-oiik^ weaker and the <*uri-ent cai'ries them ])ack toward tlie 
rack, tliey swim baek and forth a<*ross the creek, their bodit^s set obli<]U(dy to the* 
current, and their tails frequently almost touch the rack. A very little of such 
exertion soon exhausts them, and frequently they go but a ft»w feet l>efoix» being 
carried against the ra<-k, when* they die in a few minutes. 

Relation hf hrtf n K'tti/Jftr uikJ in'uimlion. — It is popularly suppose<l that the 
movement of salmon in tlie rivers is largelv det(M-min<Ml bv weather con<litions. 

•~ ft ft 

Almost, anv lisherman ean tell of a notabh* run of salmon that atn'ompanied or fol- 
lowed a south wind. ObsiMvalions ma<le durinir two vears at l>attl«' Creek fish<M'v 

•- ft » 

show that thtM'c is no relation whatever between the dire<'tion of the wind and tlie 
movements of salmon. A strong wind of any dire<*tion, however, does appariMitly 
cause them to move upstream when they have lM*eu lying in a pool for some time. 
The most notable movement at IJatth* Creek fisherv in isns was <'oinci<lent with a 
strong north wind. A lain or a slight rise in th<' wal(»r usually <*auses tlHMii to run 
upstream, but not always. There is ai)pareutly no relation whatever Ijelween 
weather conditions an<l ripening of lish. 


Tin (ilinn ntai'ii mnoL — It is not uneornmon tor lishes of the salmon familv to during the breeding seastMi. Sueh is lh<* <*ase with the Atlantic salmon, the 
various white-lishes of tin' (ireat Lakt's, ami probably with other spe<*ies, and it is 
well known that adult l*a<-itie salmons do not <*at while in fn*sh water. The Sacra- 
mento saliiKUi will often snap at bright floating obje<*ts and ean frequently bi* taken 
with the spoon while on their spawning-giounds or whih* passing u[) the river. 
Sev<»ut v-five spe<-iniens were taken in tiiiswavat ,b*llv F<'rrv <lurinir October' and 
November, IImmi. Th<*y have never been known to take Io<mI, though indigestible 
material, such as leav<*s, is sonietinu^s found in the stonuu'h. A r)-inch, mature, 
sea-run male salmon was taken at I>altU» Cn-ek tisherv in October, Is'Js. The 

ft ^ 

stomach was coniraete*! the sam<' as in the ordinary adults, but contained two small 
bits of chitinous substanee lookini; ^^omewhat like portions of the thorax of a grass- 
liopper, but may hav<' bet-n portions (»r a eiustacean. 

As thev do not <*at aftrr hMvini,^ salt water, the «liir(*st ivr <»rLrans immedialelv 
begin to shrivel up. Inmost of the *<peeiiiM'ns (»f the fall run that reach the iiea<l 
of Suisun IJav the stonia<'h and ejeea hav«* alreadv <'ontra<*ted an<l thfir walls have 
become lirm. Onlv rarelv are thrv thin .md tlaceid, a*>i it f<M)d ha<l n»cenllv l)een 


digestcKl. Th« longer the tiiiif 8iiu!e k-aving h«U w»t«r, the moiv tlio digestive 
or);an8 become eoiil ractcd. 

The figures on pngm U'li-liitl illustmU' the suecHssive changes in tUeHlinientai-y 
canal as observed in sneciniens from Mont^n'y Bay, hejtd of Snisun Bay, Sacramento 
River at Saei-auiento, and at Battle Creek, fishery. 

TJie shin. — The most ininiediat,<t eliange noticeable in the salmon afl*'v leaving 
the ocean is the great increa*te in the amonntof slime that exudes from tin' skin 
upon removal from the water. This point is of physiological interest, but iias 
nut yet been stndie<l. By the time the fish reaches the spawning-gi-oniids llie skin 
in most cases has thickened considerably, and fix^inently the scales are entirely 
embe<ld«d and invisibl<'. In the upper figure of plate I'-i the scales can \k- seen only 
where the skin has been worn off. 



Loss in weight. — Many weij|:hts and ineasiireiueuts were made in 1900 for the 
parpose of determining the loss in weight sustained ])y the salmon dnrin<i: their resi- 
dence in fresh water, but our scales prove^l somewhat inaccurate, and the data can 
not be used in detail. The loss wa^s shown to be* very larjrc, about 35 per cent. 

Weights were again taken in 1901 with accurate scales. The results are shown 
in the tables on pages 128 and li'O, which give: 

First, the length in centimeters of the specimens weighe<l, the measurements 
being ma<lc from the nostril to the last joint in the spinal column. The nostril was 
selected as a point of measurement rather than the tip of the snout l>ecause the 
snout becomes lengthened in bree<ling males. 

SliiuuK-h, pyloric aii|M-iMlaK<'<-. a"*^ iiiti'stiu** «»f twn ft-iiiaU* Miluion (Intwii t<>Miin<> m-uIo. f '.Hiiul «'f Suisuii Bay. July 
:U. im», Jiftor II fust of a «i-«-<«k or two. IK Su<*rain«'Uto, Autruht :?.». I«.««K aftt-r a •*till loii^tT fast. 

8ec(mtl, the av«Magr weights of siHn'imens of various lengths delivered at the 
cannery at IMack Diamond SeplemlH»r 5 to 11. 

Third, similar w<Mghts and averages for siH^cimens taken uimn their arrival at 
the spawning-grounds, but b4»fore they had begun spawning; als<i the iKM-centage 
of loss in weight in these siiecimens, the average weight of Black Diamond si)ecimens 


I' tiihIi-s W>' 

of th«' sHuic U'upth IwMng I'tikoii as a basis hi cac^Ii wist". The iiialcs w>Tt' weighetl at 
Battlt* Cnfk (K-tolHT •2'J, thv females at Mill Cn-ek Novi-iiiIht 15. 

Fourtli, similar weights, avcmyes, and losst's of Hjieiit lislx-s, laki'ii cither iM'fore 
OP i nutted lately afttr death. Sin-cimetis riiiiiid hetwtH'ii ()cU)Ikt 10 and Novemln'r 'M). 

In dctiTmi iiinjt the perceiitaye of loss ill the weightnf males iio aecoiiitt is taken 
of the loss of milt in spawuiiig, whieh is very sliglit, the averaKe t^rtal weijjht of the 
Bjteiit fislieH or tliosi- just arrived at Battle Onn-k beinjr eonipaitHl with tlie average 
ttiUtl weight of Bla^-k Diamond KjN'cinieiis of the saDte length. 

Crwh. OvtulH.-] 

il rvimli'. 






VI. mi 

t».Tin I 



Tho iH^rcenlHgo of loss in ripe females is determined by eomparinj^ the total 
Hveraj^<* weiglit with the a\'erage weijicht of siH»eimeiis of the same length weighed 
at Black Diamond, the latter being taken as a basis. The loss percentage as stated 
for th<* si)ent fislu^s does not inelmle tlie loss of ova. The jK^rceutage is determined 
by comparing the weight of spent fishes plus the weight of the extruded ova with 
the weight of Black Diamond fishes of tlu' same length. The weight of the extruded 
ova is determined by finding the difTen'Uce in weight between the ovaries of spent 
fishes and those of fishes just arrive<l at the spawning-gi-ounds. 

Tnhlf nf avvntiji' tnitfhts nf fmuilr stilmoti. 



Blurk I 

1m.t of 











l>er of 



( >var>'. 



1 ..« 



, (iti 

1 tiT 





1 70 






: 71 


' 72. 


. lOJKiO 





1 7H. 
. NO 

i X2. 






2, 192 
2, 775 










A venial • 

l>er of 





j Ovary. 























r,. :<»< 











"" 1 









>» i 

Tile averag<'s (»f th<* loss p4»rcentages are: For males upon their arrival at the 
s[)a\vning-grounds, H\ p<M-eent; for males at time of death, including loss of milt, 
'2*') per <'ent;.for females upon ai-rival at tin* spawning-grounds, 1-J per cent; for 
females at the time of death, not in(*luding loss of ova, l'.» per cent. The difference 
betwr»eii the loss as (h»t(M*mined in IImm) and lIMil is accoiinle<l for by there having 
biH*n many more grilse weighe<l at lh<* spa\vning-groun<ls the fornuM* year. (See 
''Two forms of males," i)agt» l*5o.) 

Under the heading ** Details of migration" (page 1-4) will be foinnl an account 
of til re(» salmon that wen* release<l in the SacranuMito near Uio Vista, after being 
w<»ighe<l and branded, and»(piently taktMi at the Government fisheries. One 
had lost 1.") per cent of its weight, another 'J") i)er<'ent, and the other '2i\ percent <luring 
the migration. 

Om* important point to be coiisid<M*<.Hl in this stu<ly of the loss in weight during 
migration is the deterioration in the valui* of the fiesh as a food. The loss of 1'2 or 
ir> or "Jo p(»r cent is entirely in nutriment. If even a very fat beef were starved two 
months, or until it ha<l lost 1>'» p<'r cent of its total weight, no one wouhl can* to eat 
of its flesh. But such is th<' coiMlition of tlie fall salmon upon thiMi' arrival at the 
upper [)ortion of the river. TIh\v have eaten nothing for over two months, and nutri- 
ment to the extent of alK)ut KJ percent of their weight has betMi absorbed, almost 
wholly from the flesh. 

It is evident, therefore, that tlie fall salmon taken at the upstream i)oints have 
but little value as food, and their capture* should Ik* [)rohibited. 

F C B li»2— 9 




Relative changes in fresh water, — Before entering the fresh water the two 
sexes are identical in ai)i>earance. With the fall run the sex can often be distin- 
guished by the external api>earance in specimens taken at the head of Suisun Bay, 
and it can always be distinguished in normal specimens by the time of their arrival at 
the spawning-grounds. Soon after entering fresh wak>r several cartilaginous teetli 
appear in the front of tlie jaws of the males, and at the same time the jaws begin to 
grow longer. By the time th<j males reach the spawning-grounds tlie jaws are much 
prolonged and liooked, and the teetli have grown to be large and solid canines; the 
body becomes deep and slab-sided, and the color usually more or less reddish. 

The principal change in the females lies in the diminution in the muscular tissue 
of the back and sides and in the distension of the abdominal walls on account of the 
development of the ova. Their color is usually more or less olive. After spawning 
the female is as thin as the male, but the jaws are not prolonged. 

The following illustrations indicate the changes better than descriptions: 

Head of male salmon taken at Sacrament^), September 5, VM\ 
HhowinK the tx^giuninfr of tlie jaw proloUKation and canino 
teeth while yet (*artila^inou8. 

Head of female salmon taken at Sacramento 
SepU*mber 5, IMN). Thin is alHo the head of 
the male in salt water. 

Two forms of )tuiles, — The illustrations on page 131 show the two forms of male 
salmon, known as adult and grilse, that are found in the headwaters. Those here 
shown were nearly of the same length, though it is v(>ry rare to find as small a speci- 
men as the upper that has the adult form, and the lower was a rather large grilse: 
(See also plat^ 13, photographs of these same specimens.) 

The differences are obvious. The grilse simply fails to develop the character- 
istics of the breeding nmle, viz, the i)rolonged and hookinl jaws, the large, hooked 
teeth, the deep, slab-sided body, and red color, and retains its salt-water appearance 
except in the loss of flesh. Grilse weigh from a half pound up, and intergrade with 
the adults both in weight and api)earance; specimens with a length of 00 centimetc^rs 
(35.5 inches) are occasionally found. I have seen two, which, from their olive color, 
could be distinguisheil as sea-run fishes, that were only 13 inches long. 

At Battle Creek fishery in 1900 the males were nearly all grilse, though there 
were almost as many of the adult males as there were females. The great prepon- 



derance of the p-ilse over the adult males and females is due to their l)eiug too small 
to l)e taken bv the nets of lawful mesh. 

The cause or that lead to the pro<luetiou of the grilse form are not known. 
Mention is made in another place in this ivport of the sexual maturing of the male 
pari-s that remain in fresh water during their first summer. It it ix)ssil)le that this 
stunts them and causes the production of grilse. If grilse were simply young indi- 
viduals that followed the adults into fresh water, we would ex^wct to find females 
among them. The two forms can not Ik? distinguishe<l except in breeding fishes. 

The dwarfed form of the female is pra<*tically unknown. Among many thousand 
specimens handled at l>attle Creek fi.shery only (Uie dwarfed specimen was found. 
This was 10 inches in total length and weighed '2 pouuils. Its ova were mature. 


Two si>t'Ut uu4l»* ^^ahnon. adult and ^rrilM* f«»nus,|found dead on the rac"k at Battle (Ye**k fishery t K*tober 2S. 191B). 
L«"nxrth fr<»nj hind«'r t*dir«* of oyo tt» \nis*' of tail, lar^rer sjie^Mnieu oin) mm., smaller 4-55 mm.; w«'i|»ht, larjjer speci- 
men. H.l«»» ifrani^' smalU-r 1.2i«» jLcram.*-. 

('omjyfinitin- stittoncnt of m'iijht ami hmjth af thv smtuf in adnjf and tjrilst* salmon of ffw samr 

hixly h*ngth. 

I^»npth from 

hinder edjre 

of eyetolia.'s*' 

of t'uudal tin 




2 5ii» 
1 . ^75 

Kht in ^rrams. 
*"^^^^*' of adult. 

i.«e5 nii» 
I. III! r>5i» 

1.575 15»» 
l.t^iii l«lit 
1,U75 1.275 
l.fu5 2«<» 
1.525 ie5 
l.N*i 75 
1.725 l.H«» 
1,:<|> ?<75 

* millii 








of snout 
of < ye 


Length from 

hinder e<l>^e 

of eve toliH--^* 

of r'audal fin 





yrht in \ 


2. 775 


• millii] 










of snout 
• >f eye 
ueters » . 


of adult. 

" mV 

1 22> 
911 » 
i 1,275 









57 _ - 


2. 475 
2. 775 




















Herm<iphr<Kiitfs. — 1 »rii inilpbli^'l I(» Mr. J. I*. HnhoCH-k, (if lh« Ciilifornia Fifih 
CoiiimiHMidu, ami t-o Mr. ("Imiiilierlaiii for two h»rnmp)ii'iKlito SHliiion. Mr. (!hain> 
borlaiii's HiK«-iiiiuii was dinooverwl I»y llio )*imwniiiK<-r<*WHt Itattlt'Ci-et^k imt<'hery in 
Doci'iiiiber, HUMIjllio other s[)i>i-iTnfii was ilisoiiverod liyMr. F. A. Cdlcswhilisclcaninn 
fish for till' canntsry at. Iflack ]>iaiiioiid, and wa.-* taken in that vi<-inity in May, liiOl. 
The mt'ompanying illustraticm n-priwents Iho Hlavk Diamond specimen. 

Th« gmiital organs nf l.lie two si)euiiiieiis are esmrritially the aamo in stnietnn'. 
Thero is l>nt one x>ai)', as in onliiiary individiiHlH, but each organ is dvvulopetl I>ar11y 
us («Htis and x>ai'tly i\n ovary. One organ in the IMack I>iamon<i si^eeinien has 
about n inches of tlio aiiterir>r portion well develo^iod a.s a testis, and iiwarly mature. 
IiuniiMUattily Iit^hiiid this arc a few ova that 
are almut as lar^e as ova ordinarily are in 
salmon lakeii in thisi>artof the river. Thp 4 or .1 inelies of tlie organ eonsists 
meivly of tlie supporting meniliraiie and 
seminal duel, with half a dozen ova devel- 
opeil in one plaee. Then follows a portion 
alH»ul -l inelies long (leve!oi)ed as t«stiti. 
TJie usual seminal duct loads i>ost«'riorly. 
'lliv ottier organ of ttiis speeimen also has 
the anterior portion develoi>ed as testis, 
while all of tlie |H)sterior i>ortion is ovarian. 
While the ova iii'c of normal sixe, the local- 
ity being considered, tlieir nunilH'r is not 
over one-fonrlli us gi-eat as wcnild bo pro- 
ducwl by n similar portion of a normal 

The liattle Creek speeimen i» similarly 
developed. Some of the ova are att^aclied, 
while others are fi-e<-, as if the lisli ha"! Ix-en 
only jiarlly riiie. All are variable in size, 
but none of normal apiH'arance are as large 
as the averagi^ ova t^tiken at ISattle t'n-ek. 
Many of the ova evidently were dead at llie 
time the lish was taken, and .s<nne of these 
were abnormally large. I uiiderslainl that 
some of the free ova were sperinati/x'd with 
le organs, but none of tlium lived. Tt is not 

lent me, and I do not know 

milt from tin' testis portion of the s 
known whether they were fertilized. 

The genital organs (uily of these si>eeimeii 
the condition of the eloa<Ml openings. 

litliifiri' iiniiilifi- '// midt'S anil ffinulefi. — lu measuring aud weighing salmon at 
Black Diamond, on August 20 and 31, lIHJi), it was noticed that the females were 
more numerous than the males. To determine whether thi.-j was merely a [teeuliarily 
of thei^ateh of those two days, 5Ir. ]■'. V. Ilnbter, the weigher for the Itlack Diamotid 
eaimery, was employed to make notes of the relative nnmlRM' of males and feuuih-s 
during the season, with the following result. 



Table Hhoiring rW«/iiv numlH'r of male a lu! female salmon taken at month of river. 


Mali-s. F».*inal*»»». 


Males. Feinali***. 

Auf;. 21 . 
Auk- 22 
Auf?. 2K 
Aui;. 21). 

















S<^pt. 1 

S'pt. 3 

Si'pt. 4 

Sept. 5 - 


Sept. 7 

Sept. s 


Ti>t«l .. 








If Si 









1 . 7>''.« 




During: OotolKM* aii<l Xovenihor, 1!mmi, Mr. Arthur SiMjj:is4»ii <*aii^ht 7«*i male and 
32 feinalo salmon on si>o<)n hooks at Jelly Ferry. 

The followinjj: is a statement of tlie relative mimlHM-of males and females taken 
in the seine at Battle ('n»ek fisherv in I'.mm): 

Table showing relative nnmln^r of salmon taken at Battle Creek fitOtery, 











Nov. 1 ( river » 





' Oct. 12.. 

Oct. 14. 
Oit. 2l>. 

Oct. 23.. 











28 i 


Nov. 19 







lo6 . 

Rehitive we'ujht, — The average weight of all salmon taken at Black Diamond 
canner}' from August 20 to Septeml)er 10,, varied daily from 21.3 |MJunds to 
23.8 pounds. As the largest 8i)eeimens taken were invariably males, it is probable 
that the average weight of the males was greater than that of the females. The 
reverse was true in the upper river, as will lx.» seen from ihe following statement 
of weights of s^ilmon taken by Mr. Sergison at Jelly Ferry: 

Table slanriutj relatire irrifjUt of male and ftniale salmon taken at Jellif Ferry, 


S»>pt.:y -. 
<>rt. 1 . 

<>ft. 2... 


31. o 

12. «» 
.. 1.5 
3 II 
3. 5 




... 4.5 






10. o 


Oct. 7 .. 
Oct. H. 
()it. W.- 
Oct. 10.. 

Oit. 13.. 
Oit. 14. 

Oi-t. If... 

ih\. 17. 
o»t IS.. 

0«t. 22.. 




.... --- 

' 4.0 

"*" "3.0 

. .. fi.o 



.'i. .5 



.1. .» 




. 14.0 



. IhiutkIs. 


Wei>:ht . i»oiin<L**. 
Miller. Fenial(>s. 

<Hrt. 24.- 

Ort. ffi 

Oit. 38 

0<'t. 27 

<Vt. 2S 

Oct. 2W 

0.t. 31 

Nov. 1 

Nt»v. 3 

Nov. « 

Nov. 10 

Nov. 12 

Nov. 13 

Nov. 15 .... 
Nov. 2r» 






4 fi 






! <vt. :t. .. 




14.0 . 


:ko : 





lit - 

»!■. .1 ... ... 


H.O 10.5 
2{.0 13.0 


lo.o 12.0 

" :'e.5 




' Oct. 7... 

12. u 

HI 1(L8 




The average weight of tlie Battle Creek specimens could not be determined, 
owing to the selection of the larger males for spawning, but it was certainly less 
than that given in tlie Jelly Ferry record. 

It will thus be seen that throughout the fall season of lOlXl there was a greater 
proportion of female salmon taken by fishermen in Suisun Ha}' and the lowei* river. 
On the other hand, tlie small males, l)eing t(X) small to 1m» taken in the reguhition net 
of the market tishermen, were greatly predominant in the headwaters. The (evidence 
here given does not indicate that more of one sex is produced than of the otlu^r. 

This point should be considered in making laws governing salmon fishing. The 
small males are not desirable for propagation, either natural or ai'tificial, an<l on 
account of their great numl)er they are a nuisance at the (Tovernment fisheri(\s. 
They are simply so much valuable food wjisted. The prc^sent law prohibits the use 
of nets that w^ould catch them, and it should be amended. As there are no small 
femah\s, the small-mesh net would not affect the supply of bre(Mling females. A 
law prohibiting the taking of small fishes is of value only wh<»n the small fishes are 
gi'owing fishes, l^ut the small salmon that come in from tlu^ ocean are not. growing 
fishes. None of the salmon ever return to salt water. Their sol(» value lies in adding 
so many pounds to the market supply or in rei)roducing their kind on the spawning- 
beds. A large fish is worth more on the markets than a small fish; but so are large 
cattle worth more on the market than .small cattle, yet a stock-raiser would never 
think of selling his fine cattle and keeping only the runts to breed from. It would 
be l>ett^r for the j^almon as a species, and then^foi-e better foi' th(* salmon industry, 
if the present minimum net-mesh were made tlie maximum. A small-mc^shed net 
does not catch so many large fishes, which would allow the larger individuals to 
reach the spawning-grounds. The salmon will certainly deteriorate in size if the 
medium and larger sizes are taken for the markets and only the smaller with a few 
of the medium allowed to breed. 


Spaivning habits. — Salmon in spawning usually take a position at the upp<M* (»nd 
of a riflfle where the current is strong and wluM-e there nro gravel and col)l)Iestones 
among which the eggs may lodge. After selc^cting tlie plac<» the female <^\trudes a 
few eggs and then moves away. The male immediately tak(*s her exa(*t position, or 
perhaps a point one or two feet downstream from it, and extrudes a small quantity 
of milt. In about iive minutes the process is repeated, the female always taking the 
position first occupied. This they continue day and night foi* over a w<H'k, usually 
nearly two weeks. I have observed salmon spawning at night, but have inner been 
able to watch one pair until spawning was complet(Hl. HraiuhMl salmon No. lU, pre- 
viously referred to, was only eight days in spawning, although souk^ <*ggs had Ixhhi 
extruded l>efore it was taken. Two weeks is the spawning time usually assigned l>y 
persons living in the vicinity of salmon streams, which is probalily about right. 

On account of the difficulty in seeing eggs under water, it has becMi impossible 
to determine the rate at which ova are deposited. Tln^ mot ions of the fish show just 
when ova are being extruded, but observation at a distance of 5 feet, with thc^ aid 
of a field glass, has failed to disclose the eggs. 

The female at irregular intervals turns over on her side and digs hov tail into 
the gravel. If the gravel is fine thei'e is often a <*onsiderable hillock thrown up, 
leaving a hole G or 8 inches deep and 2 feet across. This digging is probably not 


for the purpose of covering the egjrs, nor to make a spaee for them to lie in, but by 
the violent exercise to loosen the eggs from tlie ovaries. If the purpose were to 
cover the eggs it would be repeated every time any were d<»posited. Gravel does 
not drift so far as the eggs, and if siich were the purpose it woul<l not ]>e accom- 
plished. Besides, it is almost imimssible to cover t^ii^f^s with gravel; the eggs, l)eing 
almost as light as the water, slide away from tlu» gravel. Mon* than that, a covering 
of over an inch of even fine gravel kills them. The hillock, by forming an eddy at 
the bottom of the stream, prevents many eggs from floating away and Inking devoured 
by other fishes, but such are liable to be covered too deeply and kille<l in that wa\'. 
Some of the*tine sedinu^nt, howc^ver, may s<»tfle-on the v^f^i:^s and t4Mid to make them 
invisible to egg-eating fishes. The '*nest'' can hardly 1m^ nm<le as a place for the 
eggs to lie in, for the current always carries them Inflow it. 

Th<» presence of the other sex is not necessary to excite either to spawning 
efforts. 1 have set»n the femalt* spawning alone at Battle Oeek fishery, and other 
persons have reixjrted similar observations from otluM- places. In SeptemlH^r, 1000, I 
saw a male spawning alon<» near Sims, the female having l)een killed by a sportsman 
in order to get trout bait. Like observations have been reiK)rt4Ml by other i>ersons. 

Percenffuje of fertUizufion. — As one pair of s;ilmon dei>osits an av(»rage of t),(X)0 
eggs the increase would Ix* enormous unless there was great loss at some i>eriod. It 
is usually supposed that the greater part of this loss is due to a lack of fertilization 
of the ova. The great care necessary to secure perfect fertilization artificially has 
led fish-culturists to supix)S(i that the iKM*centage of fertilization under natural con- 
ditions must ne<*essarily l)e very low. Inartificial fertilization the ova and milt are 
mixed togeth(»r in a vessel, insuring a coating of milt or spermatized water over each 
ovum. In natiiral spawning the ova are caught in the eddies among th<» rocks, 
either near tlu» nest or within a few yards l>elow it. A few seconds aft<»r the ova are 
spawned a small quantity of milt is disseminated in the ciirrent to l)e carried against 
them. It seems very imlikely that a large percentage could lx» fertilized under such 
conditions. The following ex[>eriments throw some light on thr (pn^stion: 

To determine the i)ercentage of fertilization under as nearly natural conditions 
as }>ossible a l)ox was built 4 feet wide, 14 feet long, and b') inches diM'p, and a strong 
current of water turne<l through it. Alxmt 5 inches of gravel was put in the upi)er 
three-fifths. A pair of salmon were placed in the box <)ctol)er 2S, 1S<»7. A female 
not quite ripe was selected, in onler to allow a few days to Ix^come accustomeil to 
the place. Pickets nailed to the side prevented the fishes from jumping out. By 
Novem})er '1 they seemed to Ix* at home in the box, and their actions indicatc»d that 
they were ready to spawn. A f(*w eggs were deposite<l the next day. On the 4th 
the male died, having become almost entirely covered with fungus in the one week. 
Another was put in immediately, but the spawning was interrupted, as it requirtKl a 
day or two to get used to the place. The female died November 12, having dei)osited 
but few eggs. No cause of death could lie ascertained. Of the 51- ^^^ deposited, 
343 were killed while l^eing deposited. Of the remaining 1 ♦'»'.», 12!» or 7<> per cent were 
fertilized. In the secoml attempt 82 jK^r cent were fertilized. 

In 189S a pair of salmon were put in a ravine with simply a rack to prevent 
their going < ream. No eggs were deiH)sited. 

So far as the numl)er of eggs killed is concerned this experiment is not a fair 
test. The level floor of the box made few eildies, aiul the eggs were washed into the 
corners and killed. The percentage of fertilization would eeilainly be no greater 


than under ontin^dy natural conditions. Wo would expect the death of the male 
and the introduction of a new one to cause some eggs to Ix^ left unfertilized. 

November 18, 1807, I dug up five or six nests in ccmiparatively still watcM* where 
fishes had been seen spawning for a month. The siind and gravel were thrown into a 
screen and carefully sifted, but no eggs were found; Imt on stirring tlie gravel and 
cobblestones in a strong curriMit and setting the screen below so as to catch floating 
objects, lii eggs were secured; 1 1 w(»re alive and all fertilized, 4 al)out »l days old and 
the others about 28 days. 1 could see no indication of fertilization in the dead eggs. 

In November, 1808, in order to obtain eggs naturally spawne<l, 1 pla<*ed a scit^en 
obliquely in the wat<M' 2 or 3 f<»et below where salmon were spawning. The scr<»en 
was covered witli small cobblestones, that the eggs might lodge among them and 
be protected from spawn-eat ing fisln^s. The first trial was unsuccessful. The second 
secured 48 eggs; 30 were dead wiien found. All of the* live eggs were fertilized, tlie 
others could not Ik* t^^sted. 

The exp(»riment with the screen is not a fair test for the numb(»r killed, as the 
screen caught much gravel whicli pressed th(» eggs against the wires, and without 
doubt killed many more than would have been killed imder natural conditi(ms. The 
1807 experiment in securing eggs from the stream, when 15 j)er cent, were dea<l, was 
a fair test, but the number of eggs was to(j small to warrant delinit*^ conclusions. 
In both experiments all live eggs were feitilized.. 

In November, lOOO, 30 eggs were se(»ured from natural spawn ing-beils in Battle 
Creek, 25 of them evidently fertilized, and 1 certainly not f(M*tilized. The condition 
of the others could not 1h» satisfactorily determined, as they were killed in securing. 

At anothei* time a fish was artificially spawned in the creek on natural spawning 
beds, a screen being pbieed below to catch the eggs. A male was held in the same 
position immediately afterwards and milt expressed. The test was not ([uite fair, 
for although there was ju'obably a larger (juantity of milt than isdiscliarg(»d at once 
naturally, yet there was also a larger number of eggs. The (*ggs being cauglit by the 
screen and tliereby remaining closer to where the male* was stripped was of no 
advantage, as the life of milt in water is ample to allow it to come to rest. If they 
had been farther away it would have given time forth(» milt to l)ecome more thoroughly 
disscmiinated through the water. The eddies caused by our standing in tin* water and 
holding the fish prevented the eggs and milt from floating off in a natural manner. 
In two trials, 35 per cent were fertilized in the first, and .50 p(»r cent in the s<»c()n(l. 

These variousexperiments indicate ahighpercentagiMu natural fertilization, prob- 
ably over 80 per cent. It is significant that all live eggs found that liad b(»en spawned 
naturally, excepting on(^, wen» fertilized. The fertilization of dead eggs could not be 
determined, though they wen^ no more liable to be unfertilized than live ones. 

Morinlily atnoNtj ova. — These experiments also point Xo a high but indetlnite 
mortality from being covered by the gravel. The great(»st loss, how<n(^r, is probably 
due to spawn-eating fishes. In the middle portion of the* river, as at Hattle Cn^ek, 
when salmon are spawning, great numbei's of otlier fishes, mostly the split-tail 
(Pogoiiu'lifhijs), gather about to feed on the spawn. Fifty or more split -tails may 
often b<» seen lying a few feet downstream from a spawning salmon, and although 
each fisli may eat but few eggs, all together they pro])ably destroy a large p(»r- 
centage of the eggs spawned in the middle portion of the river. 

Trout have been t^ikcMi near the spawning platform at Battle Creek station with 
the st-omach an<l throat gorged with eggs, tlie waste from artificial spawning, and 3 



ova were found in the stomach of a trout taken in Mill Cn^ek Doe«*ml>er 1, 1001, several 
days after any artifi(*ial spawuinj^ had l>een done at the stat ion. As there were several 
salmon spawnin«i: in the ereok at that time, there is little doubt that the ejrgs were 
se<*ured from natural s[)awniiig-l>eds. Trout are adapted to catching floating objects 
ami are doubtless very dest ructive to salmon spawn where the salmon breed naturally. 

Other fishes, such as the hitch (Zy/r////V/), hardhead {Mylopliarodon)^ and sucker 
(Cdfosfoi/tu.s). hav<» not Ix^cn found to eat salmon spawn, though they probably do; 
the black-fish {Orfhofloit) has not even been found near the spawning- 1 >eds. A large 
Sacramento pike (JVj/fhorh&'iJfi.s) that had secunnl spawn from natural spawning- 
beds was taken in the river near the m<uith of l>atth» Creek in lOCKJ. 

Xofurnl rersu.s (irfifiridl propnijftfion, — Probably the most important problem 
V€*t remaining unsolved in connection with the natunil historv of the salmon is the 
efficiency of natural propagation. If we could segregate a certain number of fishes 
in a small stivam, then put a fine screen across Indow when_» they are to simwn, and 
later catch all the ah»vins and fry pnxluced, wc could solve the problem. Hut a 
small stream, su<*h as could be exiM*rimented with, is liable not to have an average 
number of fish(\s to pn\v upon the spawn and alevins, and the conditions would not 
be entirely natural. The following statement represents approximately the com- 
l)arative value of natural an<l artificial propagation: 

Pen^ntaj^e of Ums in | 



Natural Artificial 

X<»t siKiwn»Hl 


Killf^l ^*«•to^t• liut<hin^ 
Alevins JciINm! 







Total lo^^^ 



'1 10 to 'J* jmt < ♦Mit if unsiKiwutnl t'^^s aiv not remo\ ♦m1 l»y u^MUtiuinal st** tion 

'• No d»:'tiii)t** d.ita 

«'At U'ii^t .*i«ii>.T « i»nt if aU'VMiH jii«' plant»*<l 

From the foregoing it will l>c seen that the heavy loss in artificial propagation 
has iM^en in not spawning all th<* <*ggs and in planting alcvms both of which can be 
reuKMlied, as is elsewhcn* shown in this n»port. The total loss in artificial propaga- 
tion should not In* 15 per cent. 

There is a much greater loss when alevins are plante<l artificially than when 
thev hatch out naturallv 

(a) From a given numl>er of ova, as those pioduee<l by one fish, which is the 
basis of the pen'entage, there are more alevins to l)e destroye<l m the case of artifi- 
cial propagation. In natural propagation they have lx*en already largely < lest royed 
]x*fore they bfM'ome alevins, and there aie not s7 p(*r cent left to be destroyed, as in 
the case of artificial propagation. 

{b) l>ut even with a given nuinlx»r of alevins the percentage is greater in artifi 
cial plant iiiir It is not possil)le to scatter tlnMii as well in artificial phinting as in 
natural propagation. No amount of <*are will prevent their collecting in bunches, 
which has not l><*eii se<»n in natural propaL^atioii 

Something of I ht» value* of artificial propagation can 1m» learne<l from an experi- 
ment tried at Clackamas hatchery, Oiegon. In March, IHIm;, 5,<K)0 .salnuni fry 2.5 
inches long were marke<l by <*uttingofT the adi|M)se fin. The eg^A from which the 
fry were hatched were S|)awne<l at Bainl hatchery in September, 18'.>5. Mr. Hab- 



bard, siij^erinkMident of Clackamas hatcherj', who tried the experiment, reported 
that 375 of the marked fishes were taken in 1898. The smallest weighed 10 pounds, 
tlie largest 57 pounds, and the average was 27.7 ]K)unds. Besides tliese, 5 were 
taken in the Sacramento River in 1808. A few more were taken Inith in the Colum- 
bia and in the Sacramento in 1891)^ and also in 1900. The 1900 specimeiLs, however, 
may have been of those marked in the Sacramento in 1898. From those 5,()(X) fry -.5 
in<*hes long, costing less than a dollar to produce, fish weighing over 5 tons wei*e 
taken. That means tliat for every female fish strippe<l at the hat<*hery the fishermen 
should catch about 5 tons three years biter. About 4(K) of the 5,000 marked fishes 
were report<*d taken. We have no means of knowing how many came back to fresh 
water and e«cai)ed the net«, or how many were caught but not noticed. 


Qeneral effects of spawning. — Not withstanding their long journey from the ocean, 
the salmon reach their spawning-grounds in good condition. Tliey are not nearly 
so fat as when they left the ocean, but all their bruises aro received after arrival at 
the spawning-grounds. This fact has alrea<ly been noted by Evermann (Bulletin 
United Stiites Fish Commission 1890, p. 191). 


Spawned-oiit female. Battle Creek, Oftober 30, IflOO. 

As spawning progresses the abdominal walls of the female contract and she 
becomes as thin as the male. Her craudal fin is worn off to a mere stub. All fins of 
both sexes become more or less frayed, the skin wears off the sides of the tail and 
the prominent portions of the bofly, such as the edges of the jaws and bases of the 
fins. Fungus nearly always attacks the gills and the various bruised places and 
frequently destroys one or both eyes. 

It has lH*en supposed that the exertions of spawning completely exhaust the 
female and that she dies immediately u])on its completion. It would seem rather 
strange if there were just enough energy to spawn all the ova, and that with the 
extrusi<m of the last one the fish should die at once. Observaticms indicate* that 
the female has considerable energy left after spawning all of the ova, and that she 
(•ontinues on the spawning- beds for som<^ time thereafter. The injuries are 
ivceived only after most of the ova have been spawned. She probably does not 
know when the ova have all be<»n extrude<l, and her instinct comi)els her, when once 
spawning has begun, to continue the spawning efforts as long as energy lasts. The 
complete extrusion of the ova, sinc^ it is not noticed, is merely incidental. 

In 19(X), 14 spent females were taken alive on natural spawning-lnnls; 7 of them 
had extruded all ova, in one specimen there was 1 ovum yet unspawned, in two 

['ondKIOD of late 

■Dtamerrtin mnti-^. HhtiIi- rt. 

C. Fmule. «tih cjv« hill hn 

D. £ Two nulo. Krllw and adul 


there were 2 each, and in the others there were 3, 58, 0*2, and 107, respeetivel}'. 
If they died immediately after sx^awnin^ the hist ovnm we would not have found 
sucli a hirge proportion of live speeimens comph^tely spawne<l oiit; and if spawning 
completely exhausted them we wcjuld not have found them on rilHes but in the 
more quic^t water. In one instance wIkmi only one female was taken she was 
entindy spawned out and had lK»en seen spawning just befoi-e the seine was hauled. 
Of eoui'se, it is possible that the last ova were spawned just l>efore hauling the 
seine, but in any case the lish was far fi-om beiufr in a dying conditicm. (See plate 
IG and fig. A, plat*' 15.) 

Spent salmon. — A few sample field notes on the condition of silent salmon found 
dead at I^attle Creek in VM\0 are hen* given. Several hundred similar sj>ecimens 
were examined during the season. 

The following notes refer to spent femal<*s: 

Septeinl)er -M). Nearly si^iwned out. Nuiiierons ])arasites (coi>epod8) and a small patch of 
fungiis on each gill. Top of head \\'ithont fnnj^is. but with skin worn off. No fnngns on body. 
Fins and skin in j^ood condition. 

Octol)er 10. But two eggs left in Ixxly ca\nty. Gills aljout one-fourth covere<l with fungus: 
several gill parasite's. Skin worn off caudal fin and the rays alnnit half worn off. 

Octolx^r l.*». All but 1(» eggs spiwntMl. Die<t in shallow water. Caudal fin entirely worn off, 
but fish otherwise in go<Hl condition. But little fungus, (iills Imt slightly injured. 

October 2H. Entirely spawned out. excejit 2 ey^ga yet attiiche<l to ovary. Half of caudal fin 
worn off evenly; other fins in good condition, (iills one-third destroye<l. Small patches of fungus 
in varioiLs places on b<Kly. One eye blinded. 

NovemlK'r 1 . Specimens Nos. *3 and '^ from the river were of the same length. No. 2 had spawned 
all but 1)2 eggs, and No. :{ all but 442. No. ;J weighed \HH) grams more than No. 2. The skin was 
entirely worn off the caudal fin of No. 2, and the rays half worn off. Caudal fin of No. '^ in g«x>d 

The following notes refer to spent mah^s: 

Novem])er ."). Biidly st^arred. one eye blindcMl, skin worn off edges of fins and jaws. Another 
specimen, not badly scarred. l>lind in lK)th eyes, skin worn off snout and e<lges of fins. 

NovemlH»r T). Skin and fi<'sh worn off in several places bt»hind dorsal and on tail nearly to l)ack- 
bone: skin worn off (Mlges <»f fins. jaws, and the whole snout; Iwjth t'yes blinde<l: gill filaments 
half destroyed by fungus and parasites. 

November 10. ( )ne eye blinded: much scarre<l: little fungiLs. Another sp«H'imen, blind in ])oth 
eyes; skin worn off jaws and edges of fins: skin dead all over tail and caudal fin: nearly every gill 
filam«'nt with one or more parasitic cojx'P'mLs, and many sloughed off for cme-third their length. 

Disefisfs of intestine. — The intestine of the spawning salmon is frequently 
inhabite<l by tapeworms, which som(»times completely lill it and ext<'nd into the ca»ca, 
but 1 have never found tbom in the stomach. Thcv were much more abundant in 
1808 than in llMX). In addition to the tapeworms the intestine, especially posteriorly, 
is filled with a visci<l, gnnMiish ycdlow fluid. No examiimtion of this lias b<M»n made, 
but it is probably formed by the disintegration of the lining of the intestine, a 
catarrhal <les<iuamation such as has Ix^en found in the Scottish salmon. 

Funtjus. — Fungus as relate<l lo salmon dt»s<M-vcs special investigation. Nearly 
all the salmon that reach the vicinity of l>attle Creek fishery during SeptemlxM'and 
October become affected with fungus, which grows in velvety patcdies on various part.s 
of the iMKly. The points most commonly affected are the top of the head, the gills, 
fins, and eyes. Of 31 specimens noted on the racks at Battle Creek fishery during 
October IG, 17, and 18, 1900, 5 were blind in both eyes and 14 others blind in one e3^e, 
as a result of fungus. 


The following extra<*t« from my field notebook indicate the rapidity of growth 
of funguH, and show a condition rather worsen than the average, though by no means 
exceptional. The two descriptions refer to the same specimen; fii'st, on September 
30, when it w^as caught and Uigged and n^tumed to tlie creek; and second, on October 
4, when it was found deml against th<» rack: 

Septem])er 30. — Male, rijie, weight 2,8()0 grains. A notch in left pectoral, a slit in dorsal: 
caudal with a few small dead siM)ts, one worn through: 3 imrasites on left gill, 7 on right: whole top 
of head and upper edgt* of iK.H*toral fins covere<l ^ith fungus: skin i^artly worn off ^dd(*s of tail. 

Octol)er 4. — Fungus covering following iK)rtionH of fish: whole top and sides of head to below 
eyes, lower jaw, l>ac!k in front of dorsal, edges and liases of pect4>ral fins, upjier side of ventrals, a 
spot behind right iiectoral and one on l)ack 1)efore >idiiK)s(» fin. half of adipose fin. siH>t Ix^hind left 
pectoral, left side l>elow dorsal nearly to lat<^ral line and half way to a<liiM>s<> fin. l»ase of anal on 
left side, and Ix^lly In^hind ventrals. Left gill with seven streaks of two or thnn* dead filaments 
each: a little fungus on e<ich dead portion: a small pat<'h of fungus at tip of gill matting together 
the filaments of all the arches. Riglit gill with a i)atch of fungus at tip matting togi^ther filaments 
from all the arches, and another anteriorly on the inner arches. Skin of tail and most of caudal 
fin dead; some of caudal rays gone. 

There wei-e woi*se cases of fungus than the one here descrilxMl, but tliis shows 
what can grow in four days. Another siHM»imen that was in good (condition when 
tagged NovemlK^r 1 was "half covert»d with fungus" when si*en last on the 8th. 

The i)est almost disapp^^ared in I)ecenil)er. Figures I) and E of plate 15 show 
the extent to which salmon are sometimes affected with fungus. 

Gill panLsifes, — Anotlier common jiest of the salmon in fresh water is a parasitic 
copeiMHl which attaches itself to tlie gill filaments. There are usually not very many 
on one fish, but sometimes the gills an» almost destroyed by them. Plate 15 sliows 
an extreme example. Tlie gills sometimes d(»cay without being affected with fungus 
or parasites, as wjis found in a specinuiu at l^attle Creek fishery, OctolHT 7, IIKH), in 
which one-thii*d of the gill filaments were d(»ad. (See also plate 1.*}.) 

Dist'dsed ova, — In all of the females found dying during S«»pteml)er and October, 
1900, the ova were more or less diseased. Sometimes theiv were only a few addled 
and misshaiK»n ova crowd (m1 into the interstice's of the healtliy ova, but. sometimes 
almost all were addltnl. Occasionally there were a few abnormally large ova, half an 
inch or more in diameter, and in a specimen taken OctoU^r 12, IIKK), nearly all were 
in this con<lition. In anotlnn* taken alM)ut th<^ same time half w<»re of this charac- 
ter, while the normal ova had l)een spawned. In another a thinl of the ova were 
addled, and the others had absorbed water and w(mh» turgid. See fig. C, i)lat(^ 15. 

LeiKjili of life of fall salmon after reavhimj spairninfj-t/rounds. — S(4)tember 30, 
1900, numlH^nnl metallic tags were attached U) 3 male salmon, which were then 
released in the pool betwe(»n the ra<^ks at Battl<» C'lvi^k fishery; 1 of th(»se was found 
dead Octoln^r 5, having survived 5 days. Octoln^r 22, 3<*, othci-s wero tagged and 
released in the ik)o1; 27 of these were seen at various times, som(M)f th(Mn (luite fre- 
quently, up to NovcMubcr 1, and 5 of th(»ni wore found <lead within that time, the 
maximum time being 10 days. On Octoln^r 25 30 were tagg(»d and ndeas<Ml in the 
creek below tlie racks; S of thes*^ were found dead on the racks up t-o NovcmiiImm- 10, 
a pturiod of 10 days. Four were taggc^d and rel<»ased in the mouth of the cnM»k, 
about 2 miles below the fishery, on Novcmln^r 4. One of th<»se was s<'en on the 5th 
and again on the Sth, wh(»n it wjis almost deiwl, a pcM-iod of 4 <lays. November 9, 39 
were tagged and released in the rfvt»r Inflow the mouth of Battle Creek; 3 wei-e seen 
at the fishery on the 10th, 7 days aftc^rwards. 

U. U. S. F. C. 1901. 


I. R T«l1«irf«p«wi 


Altogether 12 tajrpnl lishos were seen after dyinjr, and lln^ average time that 
they lived after tagging was 11 days. The longest time was l»i days; some had 
pmhably been in the ereek a few days when tagged, though the freshest wen» selecte<l. 
Two wet*ks is a very fair rstimatr of Ihe length of life after reaching the spawning- 
gi-onnds. Branded sjieeimen No. lil, a ft^nale, lived but s days after reaching the 


The salmon of the genus Onrorliyurhus appaivntly has no instinct whatever to 
return to salt water after spawning. Worn-out spt'cimens aiv sometimes S4M*n drift- 
ing down stream and have been found as far down as Sacramento, though it is b}' 
no means certain that such have been on spawning-beds. In such castas they are 
simidy tin) weak to stem the current and, a<'conling to a Sacramento fisherman, 
"not fit to look at.'' Dead Siilmon raivly float, though thecurn*nt sometimes washes 
them along the bottom a short distance T havt^ seen ilead salmon lie for several 
days in rapids and have seen them in all stages of decay in sti-ong currents. Of the 
200 or more dea<l salmon that wen* marked and thrown over the upjyer nick at 
Battle Cit*i»k fishery in 11mm», only - were carried to the lower mck, which was a half 
mile further down stream. In small streams the water is often crreatlv eontaminat«<l 
by the ilead fish, and the stench is a great nuisance to people living in the vicinit}'. 

Thegivat variation in size of spawning siilmon, together with the (occasional pres- 
ence of certain scars, such as a broken nose, has led many people to doubt whether 
they all die after spawning once. The variation in size amounts to nothing as an 
argument, when we know that with alnuit On nlarkiMl fishes known to be of the same 
age, taken in the Columbia Hiver in ls!»s, the variation in size was from 10 to 57 
pounds. The broken nose could Ik* received at many oth(»r times than when spawning. 

It is somt»iimes thought that if a spawned-out salmon would float down stream to 
salt water it would revive, but such is not the t'ase. Humpback and <log salmon 
often spawn in small creeks and brooks that empty directly into the ocean, yet they 
die like other species. Tlu»y have be<»n seen dying and dead in braekish water. 
The investigation of the bluebaek salmon or redfish in Maho in is:»*> (see bulletin 
Uniteil States Kish Commission is'.ni, p. lHi>), when a net was pla<*ed across the 
mouth of a small stream <'ontaining about a thousand salmon, proved that that ii>\H^' 
cies has no tendency to n»turn to salt water after spawning. Lake Karluk, Kadiak 
Island, Alaska, is but about iM.) miles from the <K'ean and is a g]*eat spawning-place for 
the bluebaek s^ilmon. The outhM is shallow near the inr>uth, and if the salmon 
ever went back the Indians would be sure to see them, but they do not. In June, 
1807, the shore of the lake for miles was lined with the bone-^ of the salmon that 
had died six to eight months previously. 

The fact that all salmon of the genus Onrorhtjitrhu.^ die very shortly after 
S|)awning once can not Ix* questioned. 









Xaiuralist, Uttiied Siaivs Fish Commission Steamer A ibaiross. 



Naturalist, Uniteii States Fish Cottimission Stcauwr Aibairuss. 

The fishes foriniufi: the ]>asis of Iho followin^r rei)ort were eolleel^d in ISUS and 
1809 while studying the distribution of the fish<\s of the Saerauiento Basin. The 
colleetion was studied at Leland Stanford Junior l"nivei*sity, where siK»eial faeilities 
for study and comparison were afforded by thi* iehthyolo^ieal museum. 

The localities from wliich the colleetion was obtained represi^nl four basins, now 
distinct, though at one time probably tributary to Lake Lahonton. These basins 
are GrasslH)i)i>er Lake, Eaj^le Lake, Honey Lake, and Truckin* Uiver. 

Grassho[)[)er Lake is an alkaline pond, with no outlet, at the soutln^rn end of 
Grasshopper Plains, in Lassen County, ami contains no lislies. A siHicies of Agosia 
was found to be abundant in a spring emptying into the lake. 

Eagle Lake also has no outlet, tlie lowest point in the surrounding wat^^i-shed 
being over oo f«M»t a1)ov<» tlie surfac<» of th<' lake. Its water is slightly alkaline, 
though very cU»ar, and near tlu* shore supports thi<*k aquatic vegetation. Only two 
spi»cies of lishes were obtai!ic<l, the Eagle Lak(» whitr-lish (Rufilus (tllniceits) and 
a sucker {CJi(ismi.s((s rhntnbfrJuini) here dcs<*rib4Ml as new. A trout is known to 
inhabit the lake, but none was obtained. 

Two stivams were lishcd ifi Iloncv Lake J>asin — Willow Cri»ek and Sus*in Iliv4»r. 
The former rises at th<» lowest |)oint in tlie Eagle Lake watershed, an<l its upjx'r 
part is a rough mountain stream. It was IisIkmI about 1*) miles north of Susanville, 
when* it passes through an extensive meadow. Sus;in Kiver rises on the eastern 
slo[H» of Lassen Uutte, ainl above Susanville is a mountain torrent. Its lower part 
lies in the plains adjacent to Honey Lake and is dry during partof the year. 

Colh»ctions were ma<le in three streams of Truckee i>asin — Little Truck<»e Kiver, 
Sage II(Mi Creek, and Prosser Creek. The former, a <*onsiderable stivam, 15 to .*>() 
feet wide, witli w^vy ro<*ky bottom, drains Weblwr Lake and Indepen<lence Lake, 
and was fished a short distance below the outlet of tin* latter. Sage Hen Creek is 
tribut^iry to Little Truckee liiver. It is but a small stream, flowing tlirough a nar- 
row wo(KhMl vallev. Prosser ("reek is tributarv to Truck<»(» River and <lrains the 
tjible-land north of Truckee. It was fished near Prosser i>ar, when* it is a meadow 
stream from <> to 10 feet wide and <> inches to feet deej). 

The fish fauna of these watei-s is very limited. Thn^t^ sin^cies are describe<l as 
new. The colU»ction consists of l» native and - intnMluced sp<»ci(»s, distribute<l as 
follows: One siK»cies, Agnsni nthusfd, is common to thret* of the four basins and is 
probably to l)e foun<l in the other basin. Eagle Lake. Anoth(»r, Rutiiu^s oUcaceuSy is 
found ill all the basins except that of Grasshopi)er Plains. Chasm iates chamherlaini^ 

F.C.B.H«*-10 145 



of KaKlt^ I«iike, is iiol. fuiiiid eWwl«>rv. lloiicy I>Hk« ami TruckwUivcr lHwiii«liHve 
fi DHtivo upwncs ill coiiiitioti — Piiiil4>.sl<-us hihunlon, C<do.ihi>nux tahoensi.t, Itiililus 
olii'fuviiti, Atjtm'm ruhuxln, mid CiiltuHhiliUnijU. Siihiio in'tleuti in fotiiid in Iwtli biUiiiiK, 
but it haH Iwen iutriHliiccil into Triirkee BtiHiii, ami [xiHHihly into Ilnney Luke Itamii. 
Le.iu-hsi-uaeijrefjuin iu also known f i-oni llont-y I^akc liaisin, and Cori'ijiinux williiunstini 
and StUnio heiiahawi fmui Tru<!kce lianiii. S'.dvelinun foiittnall» haw l»eon iiitro- 
dueed into the latter luisin. 

PantOBteua lahoaton Riiltir, mw fiHvicn. 

Hi'Hii 4,.'". inleiiKth, ili'ptli '>.'>: I'yc (i in biwl: D. Hii.r II: A. 7: wales 17-f*l tolHl-12.4: to. "id 
liefiini ilorwil. Body terete, i.'aniliil iki]iiii<-1i- but little i'(>mi>reiwe(1: iiitiirorliitiil Hli){htly couves. or 
flat, vdilth of Imiie 'i.A in bittil: eye iHwtcrior. :l in xiiunt. 'J.r> in intenirliitul HjiiU'e. l.r> in ilixtanee 
botwe<'n eyt' ami npiN-r i-nd of (filliiiH'iiin^: Kiimit ctgnul to luilf of lieiiil, limailly runnilctl Ixith ver- 
tically luiil hi >ri2on tally. iiTDJtt'tiui; iH'ymiil the larffe iiionth: 4 towh of {lapilla- on uiiper lip, 4 
RiwHai-riitwMymphyaiMof lower liji. 1(1 iiiipillie in »niil)li<inertiwfriini('i>rner of niiintlitii inner i-orner 
(if liilie of lower li]!: ixlliiuun tin>u<1er lluiii iiiterorliital. tnjnul t» (liittani-e Iietwe«>n pnpilii: fontanelle 
prewiit. Imt lexH tliati Imlf wiiltli nf pupil in ii ti-ineh Mtn-ciiueUL ilorxal inxerted friini 49 to !i2 
bnndrt>(U1iH *if Uxly frtiui tip of Ninnt: vetitnils inserted under ninth ray of dimal, halfway 
U'tween tip of Miiint and ti[) »f niidille caudal rays: caudal 1.3 in hfiul, deeply enutrtnnate. not 
forked; ixvtoral 1.:i in hi.'ail: tai-iKht i)f doraatl alxiut 1.4 in beial. the Itane eipial to Mnoiit. margin 
nlifthlly eiuarKinate; ventrals l.T tt» l.S in litMul. Very dark, almost black above, abruptly iMiler 
lu'low, lower fina slij-luly dn^ky. Maxiunun length, alNiut 6 inches. 

lloM-ly reluteil tii l^tHttmtrnn yiHiTiKnin. but M-itli the fiillowinK differences, deteniiiueil by 
(^nniiariMin with siMi-iinens of that siM-cieH of the iwnie i>ize fn>in Pnivn. Utah. The Pnivo Hi>eci- 
utena have the dorsal 10 or II inntewl of It or 10. its dew.' ril>e*1 by Jordan & Kvei 


lutddk' <«uiliil 

•»■■*>■ , rr'ilHi-'IrTiJl. ^^tllw'lutlT'' 

Foniul in atmiKlauce in Snsan Riwr. and alw. in Little Tmckei- River and Prowwr (.'nvk. 
Tj-jieB (No. OUJS7. V. S. Nat. Ma»,l fnan Susan River, collected by Rutler mid fhamlxrlain. 


Oatoetomus tahoensis Gill & Jordan. 

Head 4.4 in length, o to tip of middle caudal rays; depth 4.8 in length; width of head throngh 
opercletj eiiual to its depth; eye .")..> in head. 2.5 in snout, 2.7 in interorbital sjjace, 1.7 in distance from 
eye to upj^er end of gill-oi)euing (by eye is meant the orbital opening, not the eye-ball nor socket): 
interorbital (Inme) 2.3 in top of head; width of isthmus li.H in head, a little less than distance 
between eye and gill-oiieninfr. e<inal to width of opercle, and also etinal t4) distance between 
comers of mouth. (Measurements made on a 7.3 inch 8j>ecimen.) D. 11; A. 7; scales 17-^ to 
105-16. Body rather slender, profile steep; snout blunt; mouth large, with full lips, covered 
with rather coarse i>apilla? which do not l)ecome much smaller toward margin of lips; upi>er lip 
with about six rows of much-i-rowdeil i)apillie; lower lip with two rows across sjTuphysis, and 
about 8 i>apillie in a longitudinal row through lol»es; j^)osterior margin of lower lip reiw.*hing ver- 
tical through posterior nostril. Orbitiil rim but little develoiKxl. middle ridge of .skull broad, the 
interorbital s^mce rather high and rounde<l. Insertion of dorsal in middle of body, its length 1.4 
in its height. Insertion of veutrals under fifth ray of dorsal. L«*ngth of caudal 1 to 1.1 in head, 
rather deei)ly forketl, middle ray 1.0 in longest. Anal reaching i>ast base of caudal, its height 
equaling length of cautlal; length of ventraLs e(iuaLs height of dorsiil; i)ectoral a little shorter than 
caudal. Lieast depth of caudal innluncle 2.0 in head. Lateral line complete, straight. Peritoneum 
black. Color nearly black abovt*. slightly mottled with i)ale yellowish below. 

Taken in Willow Creek. Susiin River, Little Truckee River, and Prosser Creek. Description 
based on specimens from Susan River. 

Chasmistes chamberlaini Rutter. new species. 

One young example. 1.7 inches long, and a dried head 3.3 inches long, which can not be referred 
to any hitherto descril>etl species, were obtained at Eagle Lake. Eye 7 in head, 3 in snout, 2.H in 
interorbital bone. 2 in distance from eye to upper end of gill-ojiening. Prenia.YJIlary spines form- 
ing a prominent hump, maxillary incline<l about 40 , falling far short of anterior nostril, its length 
from free end to tip of snout just e<iual to snrmt in front of nostril, 3.2 in head; lower jaw 3.5 in 
head. Interorbital (Ixme) 2 in head, considerably arched transversely; a low, shaq) longitudinal 
ridge along middle suture, showing even in the young exami^le. Na.sal spines verj* i)rominent; 
fontanelle closetl, coveretl by a thin l)one. ^lucus canals prominent, but probably intensifie<l in 
drieil si^ecimen. Lips thin, two rt)ws of papilhe on upi)er; lower incised to lia,se. lobes small, with 
scattered papilhe. (The al>ove data fn>m the drieil head.) 

Cross s^'ries of s«.-ales 1)3; D. 10; A. 7. Origin of dorsiil in middle of Ixxly, veutrals inserted 
under sixth or seventh ray of dorsiil. Pe<--toral broad, reaching two-thinls distance to ventrals. 
Ventrals sianely reaching vrnt. the outer two rays longest. Anal l«>w, when depressed reaching 
halfway to caudal. Caudal i»eilnncle l«»ug and slender. 

Hjis smaller scales thim any other s^nvies of the geniLs. The dorsid and anal are the same as 
in C. cojH i\ but it differs from that s^nvies in the broiid int»'rorbital and the jKipillose lii)s, in addi- 
tion to the small s<alt*s. The sharj* ridge nn inten>rbital als4» s^-enis t<> be a distinctive cliaracter. 

Nametl for Mr. F. M. ChamU*rlain, of the V. S. Fi.'^h Commission steamer .!//></ / 

T>i>e (No. 505SS C. S. Nat. Mils. ). ColltH^-ti-^l in Eagle Lake by Rutter and Chaml)erlain. 

I«euciscus egregius ((^irard). 

The spei^*imens h«*re nott.Nl are not tjuite so de*'p as spe<-'imens from Winnemucca, Nev., but 
otherwise can not 1k» distinguished. They have two re<l striiH*s along side, with a darker striiie 
between. Lower part of cheek yellowish, with some yellow along e<lge of Kdly. Scales in lateral 
line 55 to r»:3. D. s «ir U; A. 8 or 9. Common in Willow Cre**k and Susan River. 

Butilus ollvacous (Cojie). Kagle IaiKt W'hitr-Jish. 

This species was met ^^ith in Eagle Lake and Willow Creek, where it attains a length of 8 inches. 
Head 3.3 to 3.7 in lio«ly; depth 3.7 to 4.5: eye 4.4 to 5 in head: insertion of dorsal 0.5; J to 0.57 of 
body from snout. Scales I5-5x in (>4-><: D. S; A. ><: teeth 5-4 or 5-5. Body elongate, little com- 
pressed, little elevate<l, regularly curvt^l from <K-ciput to dorsiil, highest over tip of i^ectoral. Head 
long; mouth «»bli<iue; jaws even, the lower forming a distinct though verj' obtuse angle with lower 
profile. Premaxillary on level with lower half of pupil. Top of head .slightly i-oncave. Lateral 
line but little dtn-urv-ed. Tip of depresse<l dorsiil over fnint of anal. Caudal peduncle long, but 
little tapering, its length from anal e<iual to head l)ehind front of eye. its thickness over end of anal 
equiU to sn<mt. This species differs from Rutins In'minr in the finer scales and in haring the same 
umuber of rays in the anal that it has in the dorsal, B. bioulur haNing one fewer in the anaL 



AgoBia robiuta Rntter, new species. 

Body heavy, highest above iiisertiun of pectorals: the ventral outline cnrred almost as mneh 
as the dorsal. Head a. 8 to 4 in body; snont blunt, but little overlapping the preniaxillary and 
never extending beyond it; mouth oblique, barbels uanaJIy absent, i)resent on lU to oO per eent of 
specimens from uny one luuility. FinsHmall; D. 8; A. T; iMtctural about eqnal hi head behind nostril. 
variable; caudal moderately forked, middle rays twu-thirds length of lougent; mdimeutaiy caudal 
rays formin({ prominent heels along upper and lower edgeiiof tail; margin of anal slightly rounded, 
the anterior rays not all produced, not extending beyond posterior raya when fin is depressed. 
Lateral line nearly always incomplet«, but with scattered pores frequently extending to buse of 
candal; scales ,16 to 77, varying about 13 in any one locality. Usually two dn^hy lateral stri)>eB, the 
tipper extending from snout to caudal, the lower branching off from the nppcr behind the head 
and ending along base of anal; cheek abruptly silvery below lateral atripe; tinged with onmge 
about lower jaw, upper end of gill-opening, and at base of lower fins. 

Type (No. 505K9 U. S. Nat. Mus.). Collected in Prosser Creek by Rntt<!r and Atkinson. 

Taken in Spring Creek, Willow Creek, Sosan River, Little Truckee River, and Proeeer Creek. 

Ooregonua williamaoni Giranl. 

Abundant in streams tributary to Trnckti.' River. A '■ native whit«t-fiMh." probably tliis (<]>ecies, 
is reported from Bigler (Taboe) and Ditnner hikes in the California Fish CommiBsiou Retx>rt for 
Salmo heoaluiwi (Oill & Jordan). Lake Taliitc Trout. 

Occurs in only the Truckei^ Basin, and titkon in Little Truckee River, Sage Hen Creek, and 
Proesor Creek. The black spots of sides much larger and fewer tlian in fiiilnio irhUim, 
Salmo irideuB Uibbons. Hin'nbiiip 7Vou{. 

Readily distingninhod from the alwve by the very small and numerous black «pot«, »h well as 
by the »lM«>ace of the red blotch un inner edge of mandibles. IntTodncetl into TmckiHt Basin, and 
possibly also into Honey Lake Baaiii. It was olwerved in Susan River and Proseer Creek. 
SalvelinuB fontinalie (Mitchill). Bniok Trtntt. 

This Bi)ecioa has been introdnfiil into Prosser Creek, where Bi>ecimen» wore token. 
Cottua beldingrii Bigenmann & Eigenmann. Illob. 

Palatine teeth wanting; no prickles on skin; lateral line broken under posterior rays of dorsal, 
sometimes a few pores on caudal )>eduncle, nsnally none. Top of he^ul covered with mumte 
pimples. Dorsal spines fl to 8. dorsal nkys l.'i to 19, anal rays 11 t<) 14. 

Found in Susan River, Little Truckee River, Sage Heu Croek, and Proaaer Creek. 





By HU(;II M. smith and L. G. HARRON 

In view of the paucity of information in re^rd to the spawning habits of oat- 
fishes, and owing to the |K>ssi])le inauguration of rat-tish culture in resix)nse to a wide- 
spread demand, we think it worth while to present these observations on one of the 
mast imix)rtant members of the family. The notes may \ye taken in conjunction 
with Dr. AUwit C. Evdeshvmer's '' Observations on the breedinjjf habits of Aiaelurus 
7ieiu///f(tfJi,** published in the Anu-ncau X(tturalli<t for November, 11)01. 

On July 8, 11*02, it was observed that among a lot of 3'ellow cat-fish (Ameiurus 
7iebfiloi<ui<) from the Potomac River near Washington, which had l)een in the Fish 
Commission arjuarium since May V.K 11*02, two had jwiired and exhibited breeding 
tendencies. They had withdrawn to one end of the aquarium tank and maintained 
themselves there, the male driving away any others whi<'h approached. The other 
fish were thereupon removed and the two in cjuestion left unmolested. They were 
kept under daily observation, and their behavior furnished the principal data on 
which this pap«^r is l)ased. In the latter part of July another pair of fish in the same lot 
showed an in<'li nation to sjmwn and afforded additional information, as did also a lot 
of eggs of the same species found in a pool in the Fish Commission grounds on June 16; 
these eo;irs, which were alK)ut ready to hatch, were removed to an ac^uarium, where 
two-thirds hatrhed the same night, the others l>eing dead the next moniing. 

Xti<t'in4ik'fni/. — The aquarium in which the fish were held was 5 feet long and 16 
inches wide on th<» l)ottom and IS inches high, the po^^terior wall inclining obliquely 
l)ackward so that at the surface of X\w water the tank was 2 f<»et 4 inches wide. The 
front was of solid glass, and the sides, bottom, and back were of slate. The lx)ttom 
was covennl with gravel and a little siind to the depth of li or 2 inches. 

The nest-making, as nuxlitied by the aitificial conditions of the aquarium, con- 
sisted in removing all the stones and sand from one end and keeping the slate ])ottom 
scrupulously clean from all foreign objects, even the smallest particles of food, sedi- 
ment, etc. In moving the pebbles, which were mostly f r(»m one-half to three-fouiths 
of an inch in diameter, the fish took a vertical or slightly oblique j)osition and 
sucked a pt^bble into the mouth, usually l)eyond the lips and out of sight, then sw^am 
toward the other end of the tank and dropped it by an expulsive or blowing effort. 
Sometimes the gravels were carried only a few inches and sometimes the entire 
length of the acjuarium. I'sually the fish swam horizontally near the l>ottom when 
ctirrying a stone, but sometimes turned obli(|uely upward and dropped it from near 
the surface. l^>th fish participated in this operation. The removal of finer sediment 
waseffcH'ted bv aciuick lateml movement of the l)odv which cjuised a whirl that lifted 
and floated the ])articles }>eyond the limits of the nest. 

The jMiirof fish more |)jirticularly under consideration, during the first night they 
were in the aquarium, remov(»d all the gnivel from over a space nearly 2 feet long and 
li feet wide, upward of a gallon of stones being transferred as described. After 



the Hccond pair of iish had cleared a similar spa<»e, a pint or more of jifmvel wa.s 
scattered on the neat: the tish immediately began to remove the ston(\s, and in a few 
minutes had completely freed the nest from gnivel. The gi*avel— regjinh^d by bass 
and other tishesasdesirai)le material for the bottom of nests - may be rcMnoved l»v the 
cattish for two reasons: (1) To have a clean place for the eggs and young, so that they 
may Im^ better guarded and agitjited as hereafter descrii)ed; (:i) to provitle a smooth 
place on which to res\; and against which to rub the abdomen. 

Upward of twenty years ago the yellow cat-tish was nmch more abuiulant in 
the Potomac River than at piesent. The marshes in Piscatnway ( -n^ek were a favorite 
j)lace for the fish to spawn, and large <|uantiti(\s were theie taken each s(»ason in 
spring and summer, mostly by colored people living near the river. The fish at that 
time of year were found in shallow water occupying depressions in the muddy bottom, 
with most or all of their bodies concealed in an excavation extending latenillv from 
the rounded depression. The Hshermen easily made large cat<*hes by wading and 
thrusting their hands into the depressions. An old colored man whom we knew used 
to ref(M* to a marsh as hivj; '""meat market,'" and would often bring ashore a sackful 
of yellow cat-tish ciiught in this way. We are inclined to bc^lieve that these tish were 
brooding, l>ut we have had no opportunity of late years to examine* them critically. 

liehaivoi' of adult fiah hef<tre fqxiumJiuf. — Two days intervened l>etween the 
beginning of the and the laying of the eggs. As soon as the nest was 
made ready, the fish became ver}" (piiet. During most of the time they rested on 
the bottom, with practically no body or tin movement, except at intervals. The fish 
lay close together, often parallel, with their alxlomens just clear of the bottom, their 
weight being l>orne on the anal and ventral fins. At freciuent int(»rvals the female 
compresscHl her distended abdomen against the smooth slate l)ottom with a quivering 
or convulsive movement, the male often accompanying or following the female in 
this action, which is obviously for the purpose of loosening the eggs. 

The second pair remained on the nest from fluly IS to S<^pteml)er 1^0, when they 
were removed to make room for other species, as it was evident no eggs would be hiid. 
During this time they l)ehaved in the same wa}' as the other pair and their failure to 
spawn can not be positivel}^ accounted for, though such an out(M)me luis been the 
rule among fish retained in the Fish Commission aquarium. The enlargem<Mit of the 
alnlomcn and ripening of the eggs go on to a point when spawning seems imminent; 
the actions of the fish suggest the arrival of breeding time; l)ut no i^^^i^ are laid. 
After a few weeks the enlargement of the abdomen subsidies, and dissection has some- 
times shown a licpiefaction of the egg mass. It has been suggested that the pr(\s(Mice 
of alum in the circulating w.ater has an injurious Jistringent action on the mucous 
meml)rane of the vent, and it is a significant fact that the change from an o[xmi to a 
closed circulation, with consequent elimination of the alum filt(»r, was scmhi followed 
by the spawning of the cat-tish first mentioned — an unprecedented occurrence at 
the Fish Connnission acpiarium. The second pair of fish had been in the alum- 
filtered water for a few days, some time before the spawning season. 

Nuitihrr^ <'lutr<wU')\ and h)cuhation of e(jgH,—0\\ July 5, l)etween 10 and 11 a. m., 
the eggs were de|>osited in four sepamte agglutinated masses on the clean Axiiv bottom. 
Unfortunatelv, the fish were not under observation at this time, although th(»v W(»n» 
watched for about fifteen minutes after the extrusion of the first two lots of eggs, 
when it wjis supposed the spawning had l)een completed. The masses of eggs were 
of nearly uniform size, about 4 inches long, 2^ inches wide, and half an inch thick. 
The newly laid eggs are one-eighth of an inch in diameter, nearly transparent, and of 


a pale 3'ellow color. The nuiiil)er of egg's deposited was estimated at 2jHM). The 
incubatory |)eriod was 5 days in a mean water temix^rature of 77 F., the lowest 
temperature \ye\ng 75 and the highest So . About 12 hours intervened In^tween the 
hatching of the first and hist i^gg^^ Active movement was observed in the embryos 
4() hours after the oggs were hiid. Fully W per cent of the oggs liat<hed into normal 
frv, a few weak and deformed frv and a few unfertilized or dead eirirs l>einir notice<l. 

Gnnrth nf ytnnuj. — When the fry tii*st emerg(»d from the {^gg they were al>out 
oneH|uarter of an inch in length, and of a yellowish, tnmsjmrent color. By the sec- 
ond dav the skin of the back had begun to darken, and bv the (»nd of the foui*th dav 
the entiiv upi)er parts were unifonnly bluish black and the under side had become 
whitish. On the third day the barl>els at the angles of the mouth and the |>ectoral 
and dorsjil spines were clearly visible through the glass front of the aquarium. 

Until 6 days old they remained on the* lK)ttom in densely packed, wriggling 
mas,ses, the largest lot in the nest and sevenil smaller lots in other ])arts of the 
a(|uarium. On the sixth day they l>eg;in to ri>e vertically a few inches aUive the 
l)ottom, at fii-st falling l>3ick at once, but gradually remaining longer above the lM)t- 
tom. Bv the end of the seventh dav thev were swinuning a^'tivelv, and pnicticjiilv 
all collected in a school ju^t l>eneath the surface, whei*e they n»mained for two days. 
They then ])egan to scatter, and subsec|uently did not school. 

The relativelv large yolk-sac had nearly disapi>eared by the sixth day, when 
they began to eat finely ground Imef liver, and they were feeding ravenously by the 
eighth day. l^tween feeding times, they passed much of the time on the bottom of 
the aquarium in search of food, which they ate in an almost vertical iK)sition, hea<l 
downward; they also browsed on the sunny side of the aquarium, where there was 
a short growth of alga^. The early growth wa.s rapid, but not uniform; on the 
eleventh day their length varied from one-half to three-fourths of an inch. At the 
age of 2 months the average length was 2 inches; but after that time the growth was 
ver}' slight, and in January, ll*08, six months after hatching, the length of the sur- 
vivors was only 2} to 2i inches. The slow growth was undoubtedly due to the fact 
that the frv were retained in small troughs where the conditions w<'re ur.natural. 

C<n\ of *(j(jH and tjoHinj, — During the entire hatchi?ig [KM-iod lH)th i^rents were 
incessant in their efforts to prevent the smothering of tlu* ^gg^^. to keep them clean, 
and to guard against intruders. The ^ggi^ were kept constiintly agitated and aeratt»d 
by a gentl<» fanning motion of the lower tins, and foreign [articles, either on the 
bottom of the nest or floating near the ^gg^^. were removcMl in the mouth or by the 
fins. The most striking act in the cjireof the ^gg^^ was the sucking of the* iygg masses 
into the mouth and the blowing of them out, this l>eing re|M»ated s(»vei-al times with 
each <'luster l)efore another lot was treated. 

The male was jKirticularly active in watching for intruders, and savagely atta<*ked 
the hands of the attendant who l)rought food, and also rushed at sticks or other 
objects introduced into the aquarium. Practicjilly the entire work of dc»fense was 
assumed by the male, although the female lycjisionally ivarticipated. 

During the time the fry were on the lK)ttom the attentions of the pjirents were 
unrelaxed, and, in fact, were increascnl, for the tendencv of the different lots to Invome 
st»ttered had to Ix* corrected, and th(» dense j)acking of the young in the corners 
s(*emed to o<^casion nuich concern. The masses of frv were constantlv stirred as the 
eggs had Ijcen by a flirt of the fins, which often sent dozens of them 3 or 4 inches 
upward, to fall back on the pile. 


The vc*r\' younjf fry w<*n? also taken inti> the mouths of the parents and blown 
out. esfx»4rially those* which lM»<*anie M'paratiKl from the main lot and were found in 
the Hand and sediinent. The old Anh would take in a mouthful of try and foreign 
|)artieIeH, retain them for a moment, and expel them with some force. After the 
younj^ }M»^n tf> swim and iK^rame watt4»red, the |)arent8 continued to suck them in 
and mouth them, and, as subsequently developed, did not alwaj's blow them out. 

An interesting^ habit of the parents, more especially the male, oliserved during 
the first few <lays after hatching was the mixing and stirring of the masses of young 
by nutans of th<^ liarlnds. With th<Mr chin on the liottom, the old fish approached the 
c/>rn(»rs when* the frv were lMink(*d, and with the Invrln^ls all directed forward and 
flexed where th(»v touched the Isittom, thoroughly agitiited the ma<s of fry, bringing 
th<^ d<»eix?st individuals to th(». surface. This act wa*< usually rej^eated several times 
in (juick suca^ssion. The care of the young may l>e said to have ceastnl when they 
lN!gan to swim fn»(»ly, although lK>th jKirents (*ontinueil to show solicitude when the 
atU^ndant approa<*lied the a<|uarium from the rear. 

When 12 days old, about 1,500 of the fry were removed from the acfuarium 
to relieve crowding, and phu-ed in a hat<*hing-t rough such as is employed for salmon 
and trout. For som<» unknown (*ause, alK>ut 1,(MM) of these died during the first three 
days. The oth<»rs survived with little or no loss, and are still on hand. 

The fry which were left with their parents continued healthy, but their numlwr 
steadily de(Teas(»d. Th(»re being no way for them to escapi\ and a closely woven wire 
screen jin^venting inroads from the exterior, it was susjxH'ted that the old fish were eat- 
ing their young, though they were liberally fed at suit^ible intervals. They were kept 
un<ler <*lose observation during the day, and were seen to be fond of mouthing the 
fry, more esiM»cially the weak(»r ones -a habit which at this stage seemed unnecessary. 
They wcm-c, fre<iu(»ntly seen to follow leisurely a fry, suck it in their mouth, retain it 
for a while, and then exi)el it, sometimes only to capture it agtiin. There was no 
motive pursuit of the fry, and the tendcuicy seemed to l)e to spit them out. In one or 
two instances, however, it app(»ared that fry taken into the mouth were not lil>erate<l, 
the fe<»,ding instinct becoming pammount to the parental instinct. After all the fr}- 
which had b(»en left with their parents had disapi>eared — in al>out <> weeks after 
hatching - 18 fry from the trough wt^e placed in the iMiuarium one evening, and only 
2 of these had survived on the following morning. 

i>uring the entire period covered by these observations liver and beef were fed 
regularly to th<»- brood fish<»s, and at no time did their appetites fail. There was 
ai)j)arently no int(»rferenc(^ wMth deglutition, or closure of the (esophagus, such as 
has be(»n obs(»rv<*d in some other cat-fishes, as half-inch cubes of meat were readily 
ingested during the entire*, time the fish were under ol)servation. 

Krfrrnaf Ht',rual clnmuirrH />/*</y/////.v.— Besides the fullness of abdomen which the 
mass of eggs gives to the fenuile, there wiis in lK)th ])airs of fish under consideration 
another external feature by which the sexes could be distinguished. This was the 
shajM^ of the snout and interorbital region, which in the males were noticeably fiatter 
and broader than in the f(»mal(\s. The males in both these eases w^ere about 12 inches 
long and were an inch longer than their partners. 


By A.. K. 13H:A.RD?^I^EV, 

PmfessDr of /iioiotj^w CoioraJo S/a/r \orfnaf St'hool. 



Professor of l)iolof^y\ Colorado State Norma! School. 

The following observations were made during an investigation at the United 
States Fish Commission hatchery, Leadville, Colo., in August, 11H)2. On August 4, 
some eggs of the black-spotted trout in a numlx^r of the hatching- troughs were just 
hatching, while in others the young tish were several days old. Eiich trough was 
separated by screens into three divisions. The first division — that into which the 
water enters from the supply pipes — contained no i^)^^i>^ these having all been removed 
several days before on account of the great mortality of the young fishes hatched in 
this division of the troughs. In the second, or middle, division, the newly hatched 
fry were dying in ^considerable numJ)ers, some Jicfore leaving the ^^y[, trays. In the 
third division of these troughs, as well as in the troughs not directly fed from the 
supply pii)es, the death rate was merely nominal. 

These facts clearly' indicated that the aiuse of the mortality was directly connected 
with the water supply, which was found to l>e derived chieHy from two sources. 
The main supply pipes were feil from Kock Creek, and an auxiliary set of pipes led 
from a spring near the hatchery. Connectetl witli the main pipes wjis a branch 
leading from the third or lowest of the Evergreen lakes. Tliis was closed at the 
time, only a small (juantity of water coming through leaks around the gate at the 
head of the pi|)e. 

The water from the main piix\s wits clear, containing very little si»diment, and with 
a temi>erature of 48^ F.; that from the spring was very clear and pure, without 
sedmient, its temj^erature l>eing 43 F. There was very little sediment in the 
hatching- troughs. In this, however, microscopical examination disclosed the pres- 
ence of great numbers of a very trans|xirent hydra, which had been discovered by 
the attendants at the hatcherv a few davs l>efore, when the sun's ravs, just l>i»fore 
sunset, had fallen oblicjuely into one of the troughs. In the dim light of the 
hatchery this hydra was (juite invisible, but by placing a large mirror outside of 
the building so as to throw a l>eam of sunlight through the window, with a hand 
mirror reHecting this l>eam so as to throw it into the trough, the hydras could }>e 
plainly seen as slender, whitish threads, 1 to 2 centimeters in length and o.l5 to o.30 
millimeter in diameter, fixed l)y one end to the ])ottom or to the side of the trough, 
and bearing a crown of 5 or G long tentacles around the mouth at the free end. The 




hydi'as were found quite equally distributed through the first division of all troughs 
supplied directly from the main pi^ws. A careful count of the number on several 
square inches in different troughs gave an average of 131 hydras per square inch 
(20+ per square centimeter). Comparp-tively few were found in the middle division 
of the troughs, most of them having fixed themselves before reaching the first screen. 
Very little animal life other than hydra was found in the sediment of the troughs. 

Since no other cause for the mortality of the young fishes could be discovered, 
and as the hydi'as were exceedingly abundant and are well known to be armed with 
great numbers of dart cells or nettling cells which secrete a fluid that quickly causes 
paralysis in small crustaceans and other minute forms of animal life, it appeared 
that the injury was probably due to the hydi-as. In so far as the writer is aware, no 
injury to fishes by hydra has heretofore been known. The following ex{)eriment 
was therefore instituted to determine what injury, if any, was to be attributed to 
this cause: 

Five iKMikers, each of 250 cubic centimeters capacity, were filled with water from 
the supply pijies; in each of the first four of these was placed the sediment from 21 
s(iuare centimeters of the l)ottom of the first division of one of the hatching-troughs, 
containing about 430 hydnis; the fifth beaker was intended as a control, and contained 
water only. Five trout newly hatched and apparently in good health were taken 
from the hatching- trays and placed in each beaker. Nos. 1 and 2 were filled with 
water from the spring and were placed in running water, so that the tempemture was 
nearly constant; Nos. 3, 4, and 5 were filled from the main supply pij)es. No. 4 having 
})een kept over night in the office, and all three were set on a shelf in the hatching- 
room. At the end of the exi)eriment, Nos. 1 and 2 were at nearly the siune tem- 
perature as at the beginning, while Nos. 3, 4, and 5 had acquired the temperature of 
the hat4?hing-room. 

The following table shows the result of this exi)erimcnt: 

Tfmpemtiiro at )M>KinninK «>f exiHTi- 

Hour o( bi'ifinnliiK 

Kesult at U. I') a. in 

13^ F. 

9.i:< n. m 
1 <l*'a<l . . 

U) a. Ill 4 (load> a. in ! 4 dead 

10.:iU a. Ill Ti dead 

TemiH'niture at end of exiK-Tiiiient.. 44° F.. 


43° F. 
1 dead . . 
1 deiul . . 
4 dead . . 
r> detid . . 
44^ F.... 




4H°F 1 5H°F. 

9.20 a. m g.tKrt.m 

2 nearly dead.' 3 dead". 

3 dead ' 4 <ieiMl . . 

4 tlea«! 4 dead . . 

h deafJ . . 


WV .V)OF. 


48° F. 

9.23 a. m. 

'•One of tliese had burnt the yolk-sae in its Ntruj;gh*s. 

In this experiment 25 per cent of the trout were killed by hydms in less than 30 
minutes, (JO per cent in 45 minutes, 80 per cent in 60 minutes, and KM) per cent in 75 
minutes; those trout which were least active in the beginning of the experiment were 
the on(»s that survived longest, probably because they came in contact with a smaller 
number of stinging cells of the hydra. With the aid of a lens, the hydras could be 
seen with their mouths closely api)lied to the surface of the fish, particularly on the 
yolk-sac!; in some cuses more than a dozen hydras were seen attached to a single lish. 
Soon after the fishes were placed in the beakers most of them were seen to struggle 
violently, one of them bursting its 3'olk-sac in its struggles and dying unmediately; 
these struggles recurred at intervals, but with diminishing intensity, until death 


supervened. The tive trout in the l)eaker without hydra** were kept in the })eaker 
until the next day and were then found to he all alive and in good health. 

No other cause of injury havincr hoen dij>covered after the most careful search, 
the destructive effects of the hydras n\yon the fishes in the foregoinj^ ex[x»rirnent were 
taken as conclusive evidence that these were the cause of the unusual mortalitv of the 
trout f rj\ This fact being demonstrated, a t.-areful examination of all the sources of 
water supply to the hatchery was made. The lower of the three lakes was first 
visited. This lake is quite shallow, l)eing al)out 12 feet (4 meters) in depth in the 
deepest part; along the lK)rders there is considemhle aquatic vegetation, consisting 
of sedges and cat-tails; here the hydnis were found in immense numl)ers, clinging to 
the submerged stems and leaves as well as to the green filamentous alga which was 
growing abundantl v on the bowlders which are scattered over the l)ottom. The other 
lakes, and Rock Creek for a distance of alK)ut half a mile alK>ve the head of the supply 
pipe, as well as the spring, were examined in turn, but although very careful search 
was made, no hvdras were found in either of these waters. 

The temj^erature of the water in each of the three lakes at I foot bi>low the 
.surface was taken August S, al)out 2 p. m., and was found to Ik* as follows: Upper 
Evergreen Lake, 00- F.; Middle Evergreen Lake, (M ; Lower Evergreen Lake, 05^. 

The leaks alK>ut the head of the pipe leading from the lower lake were immediately 
stopped and no water from the lake is now entering the hatcher}'. 

The natuml causes which control the development of the different sj)eciesof hydm, 
favoring or retarding it, are as 3^et but little understood. At one period hydras may 
\ie very abundant at a given point, and soon afterwards entirely disa[)ix?ar without any 
apparent cause. They have l>een found in the vicinity of Greeley, Colo., during all 
months of the vear, sometimes in great abundance; sometimes, however, a whole vear 
has pmssed without a single one l>eing seen, although searched for most diligently. 
They oci'ur in lakes, jwuds, and marshes, usually in clear water. Warm water (tJU" 
to SO F.) a[)pears to favor their nipid multiplication, since they are usually most 
abundant in summer and early autunm; cold water does not apparently injure them, 
however, as the writer has freciuently taken vigorous individuals in the winter, 
through holes in the ice. llydnis reproduce at certain times by i'g^i^, which f>(»ttle 
to the lK)ttom and proba]»ly remain dormant through the winter, but the usual and 
most rapid mode of nuiltiplication is by budding. Little buds arise from the side of 
the parent, soon ac<|uire a mouth and tentacles like the* parent, and after a time break 
loose and lead an indejx^ndent existence. In the lake most of the hydnis examined 
were bearing from two to six buds, showing that the conditions there were favorable 
to their rapid nudtiplication. In the hatchery troughs, on the contnirv, very few 
were found bearing buds; and these were probably recent arrivals. The conditions 
within the hatchery do not, therefore, appear favorable for their increase, and it 
only remains to rid the troughs of them in the most practica])le manner. 

As the hydra is very tenacious of life and may even be cut into several pieces 
without serious injury, each piece developing the lost parts and iK'coming, in a few 
days, a complete hydra, it is not probable that it can be destroyiHl in the troughs 
without injury to the fish i-'^^f> or young fry. By removing all Q^i^g^ and fry, briskly 
scrubbing the lK)ttom and sides of the trough with a stiff brush so as to cause the 
hydi*as to loosen their hold, then <|uickly tiushing the trough into the waste-pipe, 
luost of them can l>e removed. 


Wiien the lower diviHions of the trough contain hatching eggs and fry that can 
not readily be removed witiiout injury, the first or upper division (in which nearly 
all the hydniH will have fixed themselves) may be cleaned by shutting otf the supply 
pipes, placing a t<»mporary partition between the upper and lower divisions, and, 
after a brisk scrubbing, (|uickly siphoning otf the water and floating hydras from the 
upper division of the trough. 

As the water now entering the hatcheiy is taken from liock Creek and from tlu^ 
spring, both of which are free from hydnis, it is proljable that loss from this <'ause 
will cease as soon as all the h^^dras now in the hatching troughs and supply piixjs can 
be removed. 

It was imj)ossible to find charactei's other than those of color and size by which 
to differentiate this hydra sp(»cifically from the well-known //. fnxca Linnanis. It 
differs from that common form only in ])eing of larger size and in the entire absence 
of coloration. Among the large number of individuals observed, l)oth in the troughs 
of the hatchery and at the lake, not one showed a tnice of fuscous coloration. These 
differences appi^ar to be constant, and I pro^iose the minni jxd/ u/a for the new species, 
in allusion to it^^ lack of color. It mav 1^ descrilxid as follows: 

Hydra pallida BeardHley, new Hjtei'ieH. 

CfuirdctfTit. — Bfxly cylindrical, 1 to 2 cm. in length and .15to.8<)inni. in diameter; tentacleH 5 or li, 
when fully extended two or three tinu^a as long atj the Unly; color, milk-whit<; in reflected light, 
whitihili and tran^ilncent in trannmitttHl light. 

Differs from ty])ical ILidm fu^a in Ix'ing nomewhat larger in average size and in the entire alwence 
of fuscous coloration. 

Tijpe locality. — United States fish-cultural station, Lei4<lville, Colorado. 






K. r. u. I vj— 11 11)1 




During the investigations of the aquatic resources of the Hawaiian Islands carried 
on by ua in 1901 under the direction of the Hon. George M. Bowers, United States 
Commissioner of Fish and Fisheries, very large collections of the fishes and other 
animals occurring in the waters of those islands were made. 

A detailed report, covering the entire aquatic fauna of that group of islands, is 
now in preparation, which, it is hoped, will be ready for t -ublication wuthin the year. 

Among the fishes collected are man\' species which appear to be new. Descrip- 
tions of 57 of these are given in the present paper. Illustrations of these new 
species, together with more extended notes regarding their abundance, distribution, 
habits, and commercial value, will be given in the general report to follow. 

The types of all the new species have been deposited in the United States National 
Museum, and, when possible, one or more cotypes have been donated io each of the 
following museums and institutions: Museum of Leland Stanford Junior University 
(L. S. Jr. Univ. Mus.), U. S. National Museum (U. S. N. M.), Reserve series of the 
U. S. Fish Commission (U. S. F. C), Museum of Comparative Zoology at Cambridge, 
Mass. (M. C. Z.), American Museum of Natural History, New York City (Am. Mus. 
Nat. Hist.), Academy of Natural Sciences of Philadelphia (Ac. Nat. Sii. Phila.), 
University of Indiana (Mus. Ind. Univ.), Field Columbian Maseum, Chicago (Field 
Col. Mus.), California Academy of Sciences, San Francisco (Cal. Ac. Sci.), and the 
Bernice Pauahi Bishop Museum at Honolulu (Bishop Mus.). When possible, we have 
given in this paper the nimibers which the types and cotyix\sbear on the records of 
the various museums to which they have been assigned. 

The majority of specimens here described were obtained by us in the market or 
directlv from the fishermen at Honolulu, Oahu Island. Others were obtained in the 
market or from the fishermen at Hilo, island of Hawaii; others at Kailiia, island 
of Hawaii; others on the reef at Waikiki, near Honolulu, and one at Heeia, Oahu 

Family CARCHARIID.^. The Sand Sharks. 
1. Carcharias phorcys Jonlan <fc Kveriiiaim, lU'w Hjjecies. 

Head 4.S in len^rth; depth ti.o; witltli of heail 1.75 in it.** lenjrth; depth of head 1.8; pnont a))oat 
2.2 in head; intemrbital pj>aee 2.2; sjuioe l>et\veen tip of nnoiit an<l front of numtli 2.5; width of 
mouth 2.5; eye H in interorbital Pi>aet^; intemasal si)aee l.S; leaist depth of caudal j>eduncle a little 
over 4.S; eaudal o.5 in IkkIv; pei-toral 5.75. 

Bo<ly elongate, rather robuM, tlie tail comprestse^l; h(»ad elongate, gomewhat narn)w and 
depressed; snout long and narrowly pointe<l when viewed above, the tip rounded; eyes small, their 



poHterior inaivins about mid way In^tween tip of Huout and tir»t j?ill-<ii>eninj^; nicrtitatiii}? iiit^inbrane 
well developtHl; mimth large, very convex, the antt»rior margin of mandible U'low fnuit rim of orl)it; 
teeth in upper jaw narrow, with broad baseH, compreHHe<l, serrate, not notche<l, an«l with 4 or 5 InuuI 
cuBps behind; tCHith in mandible rather long, |H)inte<l, an<l not serrate, the e<lgea nmooth; nontrils 
without flaj», inferior, and nejin»r eye than tip of nnout; interorbital npaee broa«l and convex, the 
u[>per profile of the heail rihting gradually in a nearly Htraight line to back of he^; gill-openingn of 
miwlerate length, the [K)tft<?rior over the Uwt.^ of the ix'ctoral; [)eritonemn white or pale; bo<ly very 
finely roughene<l when 8troke<l forwanl; height of lirnt dorsal lens than dei)th of IkkIv, its origin a 
little nearer tip of Hnout than r)rigin of second dorsal; origin of second dorssd nearer origin of first 
dorsal than tip of caudal, the fin small and alxjut over the anal so that the origins of the two fins are 
opposite; caudal long, with a notvh at it« tip, det^p, the lower 1o1h.> 2.25 in the length of the fin; 
pectoral with margin of fin slightly concave; ventrals small, their origins a little nearer l>ase of lower 
caudal lolie than origin of jiectoral; bac^k convexly ridgeil, broader Ixjtween the dorsals; base of caudal 
with a pit alH>ve and l)elow. 

Color in alcohol, jmle brown, the lower ]>arts ]>ale or whitish with a bn>wn streak the color <»f the 
back along side from gill-oi>ening to origin of ventral; tijw of dorsals, tnlge of caudal, and tii> of [hh*- 
toral blackish. 

This descTiption is basinl U])on the typi^, No. 50til2, V. S. N. M. (field N<». 03747), a sjKH'imen 27.5 
inches long, obtained by us at Honolulu. The collwtion contains 4 other exanii>U»s, all from Hono- 
lulu, which we take as cotypt»s. They are: No. 12715, L. 8. Jr. Univ. Mus. (field No. 03745); N(». 
12715, L. S. Jr. Univ. Miu. (field No. 0;J74<5); No. — , M. C. Z. (field No. 0:}748); and No. 1(W5, Bishop 
Mus. (fiehl No. 03749). 

We have examined 2 other si»ecimens, each alxMit 29 inches long, obtained l)y Dr. Jenkins in 1889, 
and one fuitus obtained by us at Honolulu. 

Family OPHICHTHYH)J{. The Snake Eels. 
2. Microdonophis fowleri Jonlaii Sc Kvermaiin, new si»ecieH. 

Hea<l alKUit 5 in trunk measuri^l from lip of snout to vent; tail shorter than head an<l tnmk by 
the length of the former; eye nearly in snout or 1.5 in interorbital sjuice; snout <> in head; inter- 
orbital space aUmt ().75; mouth 2.75; ])ect<»ral a little over 4.25 in head. 

Body elongiUe, cylindrical, the tiiil ta]>ering gnwlually t«> a conical horny inmit; head <'ylindrii-al 
and iwinted; snout moderately long and jKjinU'<l, slightly fiattentMl aUive, j>roji?i!tiiig over and In'ytmd 
the mandible; eye elongate, small, anterior and su[HTior, alnnit midway in length of mouth; mouth 
lather large; lips somewhat fringinl; teeth large and cjinini'-like in front of jaws, and on vomer in a 
single row; tongue small, adnate to floor of mouth; anterior nostrils in short tuln'S near tij> of snout, 
the iK)sterior with broad flaj)S on the lij>s and o]H'ning downward; interorlntal sjmce concave, eaeh 
su))nu>cular ri«lge slightly elevati^l; ()eritoneum silvery; skin iH.Tfectly smooth; liinKl with muoms 
pores, a siiries of which encircle hea<l al>ove an<l alxMit midway in it*! length; lateral line well develoiKnl, 
pores alxjut 140; origin of dorsal slightly in advance of gill-ojH.Miing or base of jH.H:toral; inrtonil small, 
the rays just alwve the middle the longi'st, fin roundt^l; <lorsal fin long and low, iti; height alHiut ^Mpial 
to length of snout; anal similar to <lor8al, its height a trifle less. 

General color, when fresh, white, renden»d somewhat sIuuUmI on upjK'r jmrtioiis by very minuU'. 
ix>ints (seen only with a giKxl lens) of gray; biick an<l up|)er surface with numerous round brown s{>ot£ 
and alx)ut 17 indistinct transverse dark brown crossbands which do not extend over tlie dorsal; the 
interspaces In'tween the spotw on the head yellow, the i>ectoral bright lemon yellow; t»nd of tail for 
about 1 inch from point bright yellow; sj>ots on margin of dorsal brown, with yelh>w iMinlers; a ImukI 
of yellow runs from under one eye backwanl, upward, across the top of hea<l, and down under the 
other eye; the transverse series of pori^s which encin;les the hc^d aboveand alnuit midway in its length, 
with bla(^k margins, and also a similar series over hea<l along tlit^ margin of mouth, and then uj>, Iwick 
of eye, over head; jK>res of lateral line without black margins. 

This spt.tcii^ is l)ased U{M>n a single siK'cunen, the tyiH', No. 50<)13, V. S. N. M. (field No. 013431), 
ar example 23 mchea long obtained by us in the market at Honolulu, July 21. 


Family MUR^XID.^. The Morays. 
3. Mursena kailuse Joniaii & Kvennann, now gpeck'S. 

Head 7 in total lenjrth; deptli 11.5; eye 14 in heatl; snout t>; interorhital 12; gape 2.75. 

Body short, stont, anil nuMlerately conipresstnl; <lii*tance fn)in tip of snout to vent le«H than that from 
vent to tip of tail l»y a distance e<jual U\ two-thinls length of head; hea<l very small an<l jK)inted; snout 
long, qua<lrate, the jaws (Hpial, the lower i'urve<l w> that the mouth doi-s not completely cloee; lips 
thin, the teeth showing; each side of upf>er jaw with a single series of unetjual, sharpish canine-like 
teeth, inside <>f which is a single deprt»ssihle fang-like tor>th near middle of side; front of median line 
with 2 long, sharp, fang-like, tlepn*s.sihle teeth; shaft of vomer with a single series of short, movable 
teeth; each side of lower jaw with a single series of uniHpial, sharp canines, those in front largest; eye 
small, midway lK*tween angle of mouth and tij» of snout; anterior nostrils each in a i>ointe<l filament 
whoee length is alM>ut half that of eye, situate<l at tip of snout just alnive lip; posterior nostrils each 
with a long filament, t»<|ual to snout in length, and situates! just al>ove anterior e<ljri* of eye; inter- 
orhital space very narrow and flat; gill-oi>ening small, nearly circular; dorsal fm very low anteriorly, 
increasing mu<-h in height on tail; anal low. 

(tround color in life, dark brown, with fine yellt)W an<l blackish spots and reticulating lini'S, the 
yellow prtnlominating on anterior part of IkmIv; end of tail dark purplish brown; c<lge of dornal and 
anal dull <lark re<l, with short pale Iwrnls lK>nlere<l with tlarker and with small jwile sp<:»t8 interspersed; 
ground color of cheek and throat yellow, with \^\q sjM»t^ l>onlere<l with black; jaw orange re<l, with 
pale black -tHlgrtnl bars; tii>s of jaws bright coral re<l; tij>s f>f nostril rdarai»nt« bright re<l. 

Color in alcohol, Iw^ly with a ground color of light irrayish brown, marki><l with fine whitish lines 
or specks, and j>rofusely covereil with numerous small, round, white si>ot>*, each o<*ellattHl with black; 
among these are s<'attere<l larger black spots and blotches; white spots smallest on back and largest on 
belly, where some areas large us eye; a broad, dark brown bar over najie, extending on side to level of 
eye; top of head and snout with fine white si>ots; side of snout with a well-<lefine<l vertical white bar 
about midway between eye an<l tip; a short white line downwanl to mouth from front of eye, an<l a 
similar longer one downwanl and backward fnim j>osterior lower angle of eye; lower jaw croese*! by 
'.i V-shaf»e<l white luirs oiK'uing forwanl and lK)nlere<l by darker; tip of jaw with 2 oblicjue white Ixirs 
separate<l by a narmw brown Hue; hist V-shape<l white l>ar extending acros? angle of mouth and fonu- 
inga large white area at ba*se of upjK^r jaw, In-hintl whirh the angle of the mouth is dark brown; inside 
of mouth mottknl brown and white; na.s;d tilaments mottltHl with brown and white; throat light bn>wn, 
with large white sfM)ts, souh* of whirh unitt* to f(»riii ol»long s[M»ts <»r lines; gill-o|>ening not surroun<led 
by dark; anal fin dark brown, «ross<'«l l>y alniut 'JS short white l>ars; i)ost(*rior jjortion of tail cn:>sse<l 
by al>out 12 distinct but somewhat irntriilar vtTticjil white bars, whirh extend u|M)n tlorsal and anal 
fins; tip of tail bn)wnish black, with 1 <»r 2 whitish s|K*tks. Only one sj>ecimen known. 

Ty[H% Xt». .VH)14, r. S. N. .M. (tield No. <K>701»), a si)ecimen V.) inches long, obtaine*! .August 9, 
1901, by Messrs. < roMsljorough and Sindo at Raihia, Hawaii. 

4. Gymnothorax vinolentus Jon Ian i"fe Kvermann, new sp<H'ies. 

Head 7.2 in total length, .'>.<> in di.*Jtan<*e froui tip c»f snout to vent; depth 14.5 in total length; eye 
14 in hi'ail; snout (J.4; gai>e 2; interorbital sji; vent a little nt»arer tip of snout than tip of tail. 

Iknly hmtr. but stout autl not gn»atly comi»n»sse<l; tail m<Mlenitely stout and compre«»ed; head 
much swollen al>ove; snout long and slender, the anterior profile iu*<*ending somewhat al>nii»tly fn»m 
inten^rbital n^ion; mouth large, extending U'vond eye a distance e<|ual to eye and snout; lower jaw 
pn^jei-ting. strongly curve<l,so that the mouth dtn»s not comi>l<'tely close; eye small, over anterior half 
of gai>e; interorbital narrow, alM»ut half greater than diameter of orbit; anterior nostril in a tul)e whoee 
length is l.iJ times eyt\ situat<*<l near tip of snout; j>osterior nostril slightly anterior to vertical at front 
n{ orbit, oval, surrounded by a narrow, raise<l, Hatteneil flap whosi.' diiuneter is twt>-thirds that of 
orbit; lips rather thin, not roveriui: the teeth; gillH>iH»ning small, its length U*ss than diameter of 
orbit. Teeth in a single s<'rii's on ea<*h side of upfH'r jaw, the jM^sterior ones short, sharp, and close- 
set; the anterior ones, al>out 12 in numU'r, slender, sharp canines of une<|ual length; inside of tht»se is 
a series of 5 or fJ long, slender, depre.*^ible canines; meilian line «»f nM>f of mouth with 2 long, sharp, 
depressible canines in fnmt, and a thinl somewhat farther Iwck; vomer with a single seriea* of short, 
blunt t<.*eth; lower jaw with a single series of rather close-set, short, backwanlly directe<l canines, 
somewhat ct)mpresse<l, inside of which ant«*riorly are 3 or 4 much longer depressible <!anines on each 



side. Origin of doriaal midway l)etween gill-opening and angle of mouth; height of dorsal 2 in 
distance from tip of Hnout to jwsterior e<lge of orbit; anal much lower than dorsjal. 

Color in alcohol, rich purplish brown or wine-color, almost uniform over entire IhmIv and hea<l; 
side of head with about 7 shallow longitudinal grooves which are darker- than ground -I'olor; im<lc»r nide 
of lower jaw yellowish white, blotched with brown; throat blotcheil with yellowish white and ]>rown; 
gill-opening rather paler than surrounding parts; iKxly everywhere witli numerous, l)ut very (»bst'ure, 
dark points, posteriorly with numerous narrow vertical dark lines api>earing as shallow gnnives in the 
skin; dorsal and anal fins uniform dark brown, not white-edged; tip of tail not white. 

The only specimen of this si>ecies which we have is the tyjK*, No. 5()H1.>, V. S. N. M. ( Held 
No. 03726), 2t) inches long, obtained by Messrs. Croldsborough and Sindo, at Kailua, Hawaii. 

5. Oymnotliorax steindachneri Jordan <.^ Kvennann, new sperics. 

Head 7.3 in length; dej)th 9.5; eye 9.5 in head; snout 5; interorbital 7.2; gape 2; tlistance from 
from tip of snout to vent U'ss than distance from vent to ti^) of tail by more than half K'ngth of heail. 

Body moderately long an<l slender, much comi)resseil; head small; snout small and jK»inte<l, the 
anterior dorsal profile concave alwve the eyw; the nape and sides of hea«l mu<'h swollen; gai>e long, 
extending far behind eye; lower jaw shorter than the upi)er, curve<l so that the mouth d<K^ not tjuite 
completely close; lips moderately thick, entirely covering the teeth in the close<l mouth; eye small, 
about midway between tip of snout and angle of mouth; teeth on sides of upi)er jaw in a single series, 
rather close-set, short, compressed, triangular canines, thost* in front s<'arcely enlarginl; vomer with a 
single row of bluntly rounde<l teeth; each side of lower jaw with a single series of rather strong, back- 
wardly directed canines, the anterior ones somewliat enlarge<l, tliose on tij) of jaw movable; anterior 
nostril in a long tube, its length about half diameter of eye, situated near tip of snout just al)ove lip; 
])08terior nostril without tulK", just al)ove anterior edge of eye; i>ores on sitles of jaws inconsi>icuous. 
Origin of dorsal fin al)Out midway l>etween gill-opening and angle of mouth, its height alniut (Mpial to 
length of snout; anal similar to soft dorsal, but much lower; tail moderately slender and pointe<l; asi^ries 
of inconspicuous ]K>res along middle of side; gill-oj>ening a long oval slit ex<*iHMling diameter of orbit. 

Color in alcohol, pale brown or whitish, sprinkltHl with ragged or <lendritic brown spots formtnl 
more or less into irregular vertical l)lotches or crossbands; margins of tins narrowly creiimy white or 
yellowish, that of the anal nmch wider; comer of mouth and space about gill-op<»ning dee]) blackish- 
brown; alxMit 5 longitudinal blackish-brown gnxjves on lower side of head; under side of lower jaw 
with 2 blackish longitudinal Hues which im^et at an acute angle under chin; thmat and iH'lly creamy 
white, with few .scattered brownish markings; sides ami toj) of head whitish, with small, sjKiringly 
scattered, irregular brownish si)0ts most numerous around and l)etween the eyes. 

This species is related to O. kidaht (S<.'hk»gel), from which it differs much in coloration, the 
[)re8ent 8j)ecies l>eing nmch paler and less reticulate<l, the angle of the mouth with mon» black, the 
gill-ojKining being surrounde<l by a broad black area (nearly or <juite absent in Mnk'if)^ an<l in having 
the white border to the dorsal tin much more distinct. 

This 8j)ecie8 is known only fn)m I^iysan (whence I>r. Steinda<'hner had 2 examples) and from 
Honolulu, where the Alhcitrosn obtaincnl 1 siKM'imen in 1S91 and tlie Fish Commij^sion 3 exami)les in VMM. 

The specimens from I^ysan which Dr. Steindachner identititni with Mnnnm tiaromnrginntn 
Riippell, and of which he gives a gcxxl tigure, evidently belong to this s]H.»(»ies. As susj^ected by I)r. 
Steindachner, the spiHiies is (juite different from ( i .jUivomartjiuaUiH^ of which siHM'ies we hav<» examincMl 
several sj^ecimens from Teilang, on the wt»st coast of Sumatra. The present sjK^cies is therefore known 
from the 2 examples which Dr. Steindachner had from Uiysiui, one s|)ecimen obtaine<l by thi* A/lMjtmtm 
at Honolulu in 1891, and 3 sjK.'cimens st^'uriHl by us at Honolulu in UK)1. 

Field No. I Lciiffth. 

03775 , 

01904 I 


131H I 










Honolulu (Albutn>^*<) 

Final dis|M>sition of s|»i>oiuu>n. 

TviM', No. riOfiKi. I'. S. N. M. 
<'otV|K?. No. 7447, L.S..Fr. Tniv. Miis. 
Cotypo. No. 2097, r. S. F. ('. 

Jlunrna Jiavomarffiiiata var., Steindachuer, Denks. Ak. WIks. Wieu, lxx, 19U0, 514, pi. vi, tig. 3 (l^iysuii); n<»t of Kuii(iell. 


6. G^ymnothorax gt>ldsborou^lii Jonlan & Kvermann, new species. 

Head nearly '^ in tnink (excliii*ive of head and tail), or 9 in total length; head and trunk about 
1.5 in tail; eye 1.75 in snout, 1.2 in interorbital space; snout 5 in head; interorbital space 7.5; mouth 2. 

Body rather compresses!, the tail gradually tapering narrowly liehind; head compressed, swollen 
above; snout pointed, the tip blunt and the sides compressed; eye rather small, a trifle nearer tip 
of snoot than comer of mouth; mouth large, snout slightly projei'iting beyond maiidible; lips rather 
fleshy and conc^ealing the teeth when the mouth is closeii; teeth in a single series in jaws, anteriorly 
large and canine-like, and the vomer with a single large, depressible fang; anterior nostrils at tip of 
snout in small tubes; posterior nostrils directly above eye in front; interorbital sp€U*e more or leas 
flattene<l like top of snout; gill-ojx'uing about equal to eye; skin smooth; head with a number of 
mucous pores; origin of dorsal a little nearer comer of mouth than gill-opening; caudal small. 

Color in alcohol, brown, covered all over l)ody except anal tin with round or roundish white spots, 
those on anterior part of Ixwiy small, very small and numerous on head, becoming larger on tnmk, 
and finally increasing very much in size on tail where they are scattered and rather far apart; reticu- 
lations around the light spots blackish V)rown upon posterior part of <lor8al fin, same color as base of 
anal; margins of anal and dorsal fins whitish; gill-opening and anus boniered with blackish brown. 
General color of Inxly in life, brown, rather pale olivaceous anteriorly, and covered all over with small 
wliite spots which are cl<«*e-set and small on hea<l where the dark color forms a network; spots sptarse 
and irregular on posterior parts, and also much larger; vent and gill-oj^ning dusky; dorsal colored 
like the body, with a bnjad white edge, growing broader behind; anal dark brown, unspotted, and 
with a bnwul pale border. 

This species is known only from the type, No. 50617, U. S. N. M. (field No. 03392), a specimen 
21 inches long, obtained by us at Honolulu. 

7. (Jymiiothorax hilonis Jordan & Evermann, new species. 

Head 8.2 in length; depth 16; eye 7 in head; snout 6; interorbital 6; gape 2.4; distance from 
tip of snout to vent 1.2 in <listance from vent to tip of tail. 

Body rather short, moderately compressed, the tail more compressed and bluntly pointed; 
head short, the nape swollen; interorbital space broa^i; a distinct median groove from near the tip 
of snout to origin of dorsal; angle of mouth posterior to eye a dLstance equal to eye's diameter; 
lower jaw but slightly curved, shorter than the upper; front of upper jaw with 3 short, bluntly 
pointed, movable teeth; side of upper jaw with a single series of short, pointed canines directed 
backward; shaft of vomer with short, blunt teeth; lower jaw on each side with a single series of 
rather long, pointed canines, longest in front and curve<l backward; anterior nostril in a long tube, 
about 2 in eye, near tip of snout just alxjve lip; ix)sterior nostril small, round, without tube, 
situated just above anterior part of eye; gill-of>ening small, its direction obliquely forward towanl 
nape; a series of 4 pores* on ea<'h side of upper jaw; similar pores on lower jaw. Origin of dorsal 
fin on nape midway l)etween gill-of»ening and middle of eye; dorsal fin well developed, its greatest 
height somewhat excee<ling length of snout; anal similar to dorsal, but lower. 

Color in alcohol, rich, velvety black al>ove, paler l)eIow where it is marble<i and reticulated with 
narrow white lines; series of p<jres on side of upper jaw and those on tip of lower, white; cheek with 
a few irregular white sfwts; gill-oj)ening whitish; side of body anteriorly with some small white specks 
and irregular whitish markings; lower jaw with larger, oblong, white cross-lines; dorsal fin rich 
brownish black, the e<ige jK>steriorly with a narrow, irregular, white border, sometimes interrupted 
by black; anal brown, with a narrow white eiige from which extend narrow intrusions of white, some 
reaching base of fin; en<l of tail with a few small white spots, the tip narrowly whit4*. 

The only known example of this species is the type. No. 50618, U. S. N. M. (field No. 04902), a 
specimen 9.5 inches long, obtiiine<i by ils at Hilo, Hawaii. 

8. Echidna zonophaea Jordan & Evermann, new species. 

Head 3 in trunk, or ().5 in total; tail longer than head and trunk by a little more than the snout; 
eye 2 in snout, 1.5 in interorbital spa<^v; snout 6; interorbital space 7.75; mouth 2.8. 

Iknly (X)mpres8e<l, the tail tapering rather narrowly posteriorly; head deep and compressed, 
pointe<l in front; snout rather long and j>ointe<i, the tip obtusely roundeti and projecting consid- 
erably beyond the mandible; eye rather small, midway between tip of mandible and comer of 



tnoutli; iiiandibh^ Hliuttin^ coinplotely, an'luMl })elow 8<> Uiat only tho anterior t(H'th toucli the 
front of tlie jaw alx>ve, though the thick Hewhy \\\yn conceal them all; te<»th molar, thoM) in front 
of jawH iH)int4Hl; anterior iioHtriln in phort tnlx«, the jM»pterior pair alK)ve the eve with a plijrhtly 
elevated margin; interorbital Hi>aee ccmvex; top of head more or lew swollen or wmvex in i»rofile; 
gill-opening 1.67 in eye; »kin wiiooth; head with a few ixjn»s; origin of dorsal lH*ginning at last 
fourth of siwct^ l)etween corner of mouth an<l gill-oiH'ning; c^uidal nmall. 

Color in alcohol, grayinh white, the iKxly an<l tail <to?*h(h1 by about 25 bn)ad rich bn>wn })andfl, 
extending u{K)n the dorsal and anal lins; dark lmndi:< anteriorly bn>adt»Ht alK)ve and not m(H'ting acrow 
l)elly, their width alK>ut ecpial to the tlistance from tip <)f snout to middle of eye; first brown Iwimi 
through eye, second acrosH nain;, the fourth across gill-ojK*ning; gray Imnds of ground color antt»ri- 
(►rly brojid, and wi<lening nmch u{Km In'lly; iK)steriorly the gray bands are narrower and l)etter 
defincnl, esp<H'ially on the tins, their width sciira»ly gR»ater than half that<>f the bn^wn bands; tip of tail 
very narrowly white; l>ody anteriorly, especially within the gray 1>ands, j»rofuHely covere<l with numer- 
OU8 small, n)undish, blatrk si>ecks, less num(ToUBand more soattere<l jMysteriorly; no black 8iK)tHon head; 
angle of mouth black, with a small white bl(>tch inune<liately in front on lower jaw, continucN.l across 
under jaw la* a broad whitish l>and; side of head with alxmt 4 or 5 narrow blackish lint»s lH»twi»en mouth 
and gill-o{>ening; n*gion of gill-ojH»ning marbletl with dark brown and whitish, the oiM'ning <lark. 

One example (No. 0.^545) had much yellow on the head and lH*tw<HMi the brown zones. 

This H|K»cieH is known from the tyjH^ and .'{ coty|K»s, all obtaimnl by ns at Ilonolnlu. 





Fiiml (lbiH)^iti<iii of h|mmmiiu'u. 


TviM-. No..'iOt;21. r. S. N. M. 
('oty|M', No. Lti'is. r. S. F. ('. 

17 <lo <'(»typr. No. TIls. L. S. .Ir. Tnlv. Miis. 




CotyiH'. No. :i9i..\ Field Tol. Miis. 

Family MYCT0PIIII).4':. The Lantern -fishes. 
9. Bhiiioscopelus oceanicus Jordan iV: Kvermann, n(>w s{>e<'ie.'i. 

Head WJy in length; depth 4.1; eye 2.5 in head; snout very short, alxMit <>; interorbital ,*J.5; 
D. about 12; A. alK)ut IS; scid<»s 2-:;5-.S. 

Hody stnmgly compresscMl, {wirticularly jK>steriorly, wIhtc it tajK^rs into the long, slender inmdal 
j>eduncle; head exciHiling de{>th (»f Ixnly; mouth large, somewhat oblirpie, tlu' jaws e<iual, the max- 
illary reaching In^yond the orbit, its |K)sterior <'nd clul»-sha{H'<l: eye large; anterior profile rather 
evenly convex from tip of sn(»ut to nape; teeth dillicnlt to make out, but a single row of minute ones 
can l)e seen on the e<lge of t»ach jaw, the exterior gnuuilar or short, villiform striiK', if it exists, U'ing 
invisible even with thcwiid of a good lens; tin^th on vomer and edges < if palatines more distinct than 
those on jaws, and forming a broader lim» as if there wen* 2 or .*{ rows; no granular fuitches visible on 
disk of ))alatine Ixuie; an elevated acute mesial line si'jKirating one* nasal prominenc<^ from th<» other; 
intemrbital space (hhiv^'x, roun<l(Ml; pn'operde nearly vertical, slojiing slightly backwanl from alK)ve 
«lownwanl; scahw larg<», undulated and very irregularly and sparingly tootiie<l or cn*nat<', and having 
about .'i Iwisal furrows; s<'ales of latenil line conspicuous and mon» jK^rsistent; 7 pliotophon*s along Uise 
of anal, 5 along lower <Mlgc» of caudal pcvlimcle, 2 at ba.Me of caudal, 1 on middle of side alwive last anal 
photophon% 4 on each side of belly lM'twe«'n vcntrals and origin of anal fin, 5 lM'tw«*cn has** of ventral 
and gill-oiHMiing, 1 on side above b;t«e of ventnil, a row of ."> upwanl and bju-kward from fnmt of anal, 
1 abov«^ and 1 Ih*1ow bas<» of jMftoral, an<l 1 on lower anterior portion of opercle: origin of dorsjil 
sonu'what U'liind biL*H» oi ventmls, the iM>sterior rays, tog^'tluT with those of anal, dividi'd to the ba.'je; 
no spine at base of <iiudal. 

('oh»r in alcohol, imiform brownish, the scak^, esjKM'ially on nnd<lle of sid«', metallic stei'l blue; 
top of h<*ad brownish; siile of head bluish; photophorcs black with silvery center; fins dusky whitish. 

This siK*cit»s wjifi re<'orde<l by Fowler from "near the Sandwich Ishuuls," as Ixltimnnitprhis mrns- 
cans ( Uichanlson), tin* record lH»ing biU<ed ujxni 4 specimens (Nos. 7W72 to 7i>75) coll(>rted by Dr. \V. II. 
Jones, and now in th<^ Philadelphia Academy. During tlu^ Agassiz South l*acifi<* I'xpedition of the 
AUhiIvohh in ISlHJ-llXK), 2 examples of thissiKHn(»s wen» taken in the surface towing net at S p. ni., Sep- 
tembers, lstM.),at latitu<lc 1()°57'X., longitu<h» l:i7° .'tr/ W., southeast of the Hawaiian IsUuhIs. These 


2 specinjens are doubt U'sh identical with thone reconled by Mr. Fowler, and are apparently dirtinct 
from H. ntriiMutnA^ the ty{>e of which came fnim l>etween St. Helena and A«*enHion Islands and others 
from l)etween Australia an«l New Zealan<l. Thev are near R. muhea' Lutken, fn>m which thev seem 
to differ in the blunter snout, the more slender tail, and in having the ix>sterolateral photophore 
8<^)mewhat before the adiposi* fm. 

Tyi)e, No. 5()(i22, U. S. N. M. (field No. 05Sa5), !.:'» inches lonjr, collected by the AltKUrftM at 
S p. nl., Septoml)er S, 1809, at the surface at 137'' [W \V., 10° 57' N. ; cotyj)e, No. 27:]6, \\ S. F. C, same 
size, <*ollcrte<l at same time an<l phuv. 
Rhino*co)Hlu.^ (^nutntns, Fowh-r. I^oc. Ac Nat. Sci. Thilrt. lyiK). I'JS (nmr tho Siindwich lslnii(l.«<): not of RichanlHoii. 

Family SYXGNATHID.^. The Pipe-fishes and Sea-horses. 
10. Hippocampus fisheri Jonlan i^ Kvermann, new species. 

Eye 2.S in snout; snout 2 in heail; D. 18, on 4 rinps; A. 4; P. 15; rinjrs 12 i IM. 

Tail longer than heail and trunk; tnmk rather di»ep, compresse<l, its width 1.7 in depth; eye 
small, e<jual to interorbital width; interorbital spa<'e concave; ^11-openinp small, high; spines on head 
ami IxMly rather hich, shar|); 2 rinp? on trunk lietween each pair of larjrer spin(»s; tail with 3 ringR 
l)etwi»en each i»air of larger sj)ines; cf>ronet well develof>e<l, with 5 spines; spines over eye blunt; base 
of dorsal alxuit e<jual to snout; anal small, lonjr; j>e<'tonil broad, rays rather long. 

C<jlor in life, tnmk Ix^low middle row of rin<rs yellowish jrolden, al)ove middle row blackish brown 
on orange ground; knobs orange; lower |K)rtion of knobs on S to 11 rings sjxjtte^l with dark brown; 
side and toj) of tail same as l>a<'k of tnmk; ventral si<le |>:de dirty orange; head, crown and snout dirty 
dark brown; an orangt» band across snout and one l>ef ore eyes; pale brownish golden over gills; chin 
orange; iris yellowish golden with 8 nnldish streaks radiating fnmi pupil; tins i)ale; a red spot l)efore 
each eye at each si<le of preorbitiil spine. 

Color in alcohol, {wile bn)wn, up{x*r surface with dark l)rowii marblings; side with small roundish 
dark siK)t8. 

The above description is from the type. No. 50025, T. S. N. M. (tield No. 0.'^*^5), a specimen 2.0 
inches long, obtaintNl at Kailua, Hawaii, where the s|M.'cies was new to the nativt»s. We have 5 other 
examples, ea<*h alxjut '.\ inrlu'.< long, taken from the stomach of a dolphiji {Cortjjthana sp.) which 
was captunnl at llilo, July IS, 1901. 

When fresh. No. 0;{507. a male, wa** i)ink or pale cardinal ahmg ami near the keels; plates on l>ack 
ami alK)ve mi<l<lle row of knobs on si<le mottk»<l blackish on pah* re<l ground; plates Inflow middle row 
of knot >s and on l>i*lly |M»n'clain white; cgg-{>ouch uniform pale canlinal-n^l, paler than rest of iKwly; 
tail same pink or pale cardinal, mottli*<l with bhu* blotches; top of head and crown blackish on 
pale re<l; clu^k, jaw, and snout pink. Sonic cxam{>les had ventral side of tail and portion behind 
fourth prominent spinr of tail uniform pale canlinal-nMl. 

This s|>e<'i«*s is name^l for Mr. Walter V. Fisher, of Stanfonl I'nivi'rsity. 

We have the following si>e<Mmens: 

V I^'imtli. I/H-alitv. FMiml tlisfMisition of >fH'(>iini'ii. 1 

_ _ _l 


(tKV* '2.t\ Kailim.... Typt'. No..=i(W2:». T.S. N'.M. | 

(CfiJiT Hilo rotviM'. No. 74.'i<». L. S. .Ir. Tniv. Niiis. 

«ln .... CotViM'. No. JTdiJ. r. S. F.r. 

Mo f'otyiM-. No. ;^.»l«.. Firl«l Col. Mu.«. 

<lo <'otv|>«». No. I«;s7, Hi<i|io|» .Mils. 

<lo CotyiM'. No. . .M. ('. Z. 

11. Hippocampus hilonis Ionian «^ Kvcnnann, new sfn^i'ies. 

f>e al)out 4 in sn<>ut; snout 2 in head: D. 16, on :; rings; rings 12 : lin. Tail a little longer than 
heail and trunk; tnmk rather dee]>, c<Mnprt»ss<'<l, its width 2 in depth; eye small, e<jual to interorbital 
width, which isconcave, broader i>osteriorIy; gill-o|»ening high, rather large; spine^on hea<i and bo<ly 
very blunt, rounde<l, or obsolete, though fonning knobs of ujore or le><s e<|ual size along tail; coronet 
with rounde<l knol>s, In'fore which is a short ki»el or trenchant ridge; Ixk^eof dorsal alx>utl..'i5 in snout. 

Color in alcohol, dark or blackish brown, more or le*< uniform. 


This species is known to us only from the example desorilied alx)ve. It is closely related to the 
Japanese Hippocamjnts ateirimiiA Jordan & Snyder, but on comparison with the tyj>e of that 8f)ecie8 
was found to differ in the presence of the keel on the top of the head and in other minor characters. 
It is also close to //. ringniJ*, 

Type, No. 50626, U. S. N. M. (field No. 03832), a specimen 6 inches long, presented to us by Mr. 
A. M. Wilson, of Hilo, Hawaii, where he obtained the specimen. 

Family ATHERIXID^. The Silversides. 
12. Atherina insularuzn Jonlan <& Evennann, new species. 

Head 4 in length; depth 4.75; eye 3 in hea^l; snout 4; interorbital 2.8; maxillary 2.5; mandible 2.2; 
I), vi-i, 11; A. 17; scales 46, 6 rows from anterior lyase of anal upward and forwan.1 to spinous dorsal. 

Bfxly oblong, compressed; head triangular, the sides compressed, top flat; mouth large, oblique, 
maxillary reac!hing front of pupil, lower jaw included; teeth in rather broad villiform Imnds on jaws, 
vomer, and palatines; interorbital sjMice very broa<.l and flat; snout broad, truncate; origin of spinous 
dorsal slightly i)osterior to vertical at vent, slightly nearer" tip of snout than base of cau<lal; longest 
dorsal spine about 2.4 in head, rt»aching nearly to vertical at front of anal; distance l)etween spinous 
and soft dorsals equal to distance from tip of snout to middle of pupil; edge of soft dorsal concave, 
anterior rays somewhat produced, their length 1.9 in head; last dorsal ray about om»-half longer than 
one preceding; base of soft dorsal 1.8 in head; origin of anal considerably in a<lvance of that of soft 
dorsal, the fins similar, anterior rays about 1.7 in head, base of anal 1.3 in head; caudal widely forked, 
the lol)es equal; ventral short, barely reaching vent; i)ectoral short, broad, and slightly falcate, its 
length about 1.4 in head. Scales large, thin, an<ldeep, 19 in front of spinous dorsal, 6 rows between 
the dorsals and 9 on mcnlian line of caudal iK^luncle. 

Color when fresh, clear olive grivn with darker edges to scales; lateral stripe steel blue above, 
fading into the silvery belly; fins uncolortnl. 

Color in alcohol, olivaceous alH)ve, silvery on sides and l>elow; scales of liack and upper part of 
side with numerous small round coffee-brown spe<'ks, disiK)sed chiefly on the edges; median line <»' 
back with a darkish striin.*; middle -of side with a broad silvery band, plumbeous above, especially 
anteriorly, more silvery lx»low; t^)p of head and snout with numenms dark brownish or black specks; 
side of head silvery, oj^nrle somewhat dusky, sides and tip of lower jaw dusky; dorsals and caudal 
somewliat dusky, other fins pale; pe<'toral without dark tip. 

This small fish is common inside the reef in shallow bays everywhere in the Hawaiian Islands. 
Many indi^^duals were seen off the wharf at I^haimi on Maui. Our colle(!tions of 1901 contain 20 
specimens from Kailua, fn)m 1.5 to 3.5 ijiches long; 43 from Hilo, 1.5 to 2.25 inches long; and 1 from 
Honolulu, 2.25 inches in length. Numerous si)ecimens were obtained by the AifnitroM at Honolulu in 
1902, 1 of which is taken tw our tyiw an<l 3 others a« coty|K»8. 

Tyi>e, No. 50819, U. S. N. M., 4.25 inches long, obtained by the Allxitrosn at Honolulu. Cotypes 
No. 2741, U. S. F. C, 3.9 incht* long; No. 2302, -Vm. Mus. Nat. Hist., 3.9 inches long; and No. 4003, 
Field C^l. Mus., 3.5 inches long, all (!ollecte<l at Honolulu by the .l/6a/r(ww. 

Family HOLOCENTRn)/E. The Squirrel-fishes. 

13. Myripristis bemdti Jordan <& Kvermann, new species. 

Head 2.8 in length; depth 2.4; eye 2.7 in heatl; snout 4.7; maxillary 1.7; interorbital 4.9; 1). x-i, 
16; A. IV, 14; P. i, 14; V. i, 7; scales 4-,32-7. 

Body elcmgate, (U»ep, compressi^l, its greatest depth at liase of ventral; head large, compressiHl, its 
depth less than its length; snout short, blunt, convex, ita width al)out twitre its length; ujuwr profile 
of head straight from alK)ve nostril to oit'iput; eye large, high, it« diameter a little less than jM>sterior 
part of head, and its upper rim hanlly im{)inging upon upi)er profile of head; mouth very large, 
oblique; mandible slightly projecting, the maxillary not reaching iK)sterior margin t)f eye; distal 
expanded extremity of maxillary 1.7 in eye; several enlargwl, blunt t<»eth on outer front edgt»s of jaw 
and sides of mandible; teeth in jaws line, in broa<l bands, also on vomer and palatines; tongue thick, 
pointed, and free in front; sul)orbital rim narrow, finely si'rraU^; lower pixsterior margin of maxillary 
with blunt denticulations; lips rather thick ami fleshy; nostrils dosi* t^gt»ther, iH)sterior very large, 
close to front rim of orbit; bones of head all finely serrate; opercle with well-developed spine; gill- 



opening large, filaments rather large; gillrakers long, fine, longest longer than longest gill-filaments; 
pseudobranchiie very large, outtT |)ortions free for half their length; dorsal spines slender, first 3.4 in 
head, second 2.6, third 2.2, fourth 2.2, tenth 6.4, and last 3.5; anterior <lorsal rays elevate<i, produced 
into a jx)int, first 1.8, sei'ond 1.7, and last 8; first and second anal spines short, third 2.6, and fourth 
2.8; 8<jft anal similar to soft dorsal, anterior ray 1.75, thinl 1.8, and last 6; caudal forked, lobes pointed, 
1.2; pectoral rather small, p^jinted, 1.4; ventral 1.6, reaching .65 distance to anus; caudal peduncle 
elongate, compres.sed, 2.2 in hea<l, its depth 3.25; .sc^ales large, ctenoid, deep; lateral line slightly 
arched, running obliquely down on side along upj)er part of <.^udal {>eiluncle; 4 rudimentary caudal 
rays above and Inflow, slender, sharp-pointed, and gratiuate<l. 

Color in life (No. 0)J370). deep red, with silvery luster; no stripes on side, a bloo<l-red band across 
gill-oi)ening and base of |)ectoral; fins <leep reil, without white wlgings, <listal half of spinous dorsal 
shading into orange. 

Color in alcohol, pale straw-color, fins plain and paler; upj)er margin of opercle blackish, and axil 
of pectoral l)lack; anterior margins of soft dorsal and anal whitish. 

We take j)leasure in naming this species for Mr. Iakhs E. Berndt, superinten<lent of the Honolulu 
market. Descril>e<l from an example (No. (K{^U6) taken at Honolulu, where the species is rather 
common. Our collections <'ontain the following si)ecimens: 


I Length Ltx-alily. 

Final dijip<j»<ition af 







" 1 








Honolulu. Typo. No. 50627. U. S. N. M. 

do Cotype, No. 16818. Bifhop Mus. 

do Cotvpe, No. . M. ('. Z. 

do.... Cotype. No. 2701.1'. S. F.C. 

do Cotvpe. No. 2282. Am. Mils. 

Nat. Hist. 

do ... . Cotvpe, No. 1489, Cal. Ac. Sci. 

... .do ... . Cotype, No. f>0rJ'2S, r. S. N. M. 
do Cotype, No. 9801. Miu*. Ind. 







Leng^th. I>M'ality. 

Final disposition of 






Honolulu. Cotype, No. 3»47. Field 

Col. yius. . 

do.... Cotvpe, No. 24212. Ac. ' 

Nat. Sci. Phila 

do Cotype, No. 7451, L. S. Jr. 

Univ. Mua. 



do ... 

do . . . 

do ... 

14. Myripristis chryseres Jordan «?t Kvermann, new species. 

Head 2.7o in length; depth 2.5; eye 2.4 in hea<l; snout 5.5; maxillary 1.9; mandible 1.8; inter- 
orbital 5; D. x-i, 14; A. iv, 12; scales 4-34-fi. 

B<xly short, stout, and compresse<l; dorsal profile evenly convex from tip of snout to origin of soft 
dorsal; ventral outline nearly straight to origin of anal l> is tH:|ually oblique with that of 
soft dorsiil; caudal |>e<luncle short but slender, and not greatly comj>resse<l, its length from l>ase of last 
dorsal ray to first short spinous caudal niy 1.3 in eye, its least wi<lth alx>ut 3 in its least depth which is 
1.8 in eye; hea<l heavy, short; mouth mcxierately large, the gaj)e in close<l mouth reaching vertical of 
mirldle of eye; maxillary very broad, triangular, reaching nearly to vertical of i)Osterior line of eye, 
with a bn)ad, curved supplemental l)order; surface of maxillary roughly striate<l, anterior e<lge near 
the angle strongly dentate; lower jaw .«trong, somewhat proje<*ting, the tip with 2 rounde«l rough 
prominences fitting into a distinct notch in upper jaw; teeth short, in narrow villifomi l>ands in jaws 
and on palatines, a small patch on vomer, none on tongue; eye very large, orbit excee<ling postocular 
j>art of head; lower eilge of eye on level with axis of l)ody; snout short, 2 in orbit; interorbital spa<*e 
nearly flat, strongly rugose; 2 long ridges from preorbitals to na[)e; outside of these a short ridge 
l)eginning alxjve fnjnt of pupil, extending l>ackward and branching upon nape; supraocular ridge 
spinesi'ent posteriorly; sulxirbital narrow, strongly dentate below, up|)er edge in front somewhat 
roughene<l; ojien-ular l)ones all strongly toothed; o|H'rcular spine short ami ol)Scure (stronger in most 
of the cotypes); dorsal spines slender, fifth longest an<l strongi*st, its length 2.5 in head; first dorsal 
spine somewhat j>osterior to base of {sectoral, its length 2 in eye, spines gradually shorter from fifth; 
spa<'e l>etween <lorsals very short, alnrnt e<jual to length of tenth spine; dors:il rays long, length of 
longest a little greater than orbit, last equal to pupil; first anal spine very short, se<x)nd short and 
triangular, its length aVxmt 1.5 in pu])il; thinl anal spine long, strong and straight, longer than fourth, 
its length equal to diameter of orlnt; fourth anal spine slender, it« length 1.3 in orbit; anal rays longer 
than th<jse of dorsal; <'au<lal widely forke*!, IoIk's equal, their lentrth 1.5 in head; jXM'toral long and 
narrow, its length 1.4 in hea<l, the tip reaching past tips of ventrals; ventrals slender, pointetl, nearly 
reaching vent and nearly as long as ]>e4.'toraL 



Scales fliTialler than in ^f. rmtrdjan, number in lateral line 34 in type, 35 to 38 in Pome of the <H)type8; 
0caleH strongly dentate, an<l striate near the e<ijfe8; a strongly dentate humeral sirale. 

Color in life, bright scarlet, ctmters of the scales paler; a blat^kish-red Imr behind, and on edge of, 
opercle, continued a<« red (not black) into the axil; first «iorsal golden, with rtnl basal blotches on mem- 
branes; second dorsal golden, with crimson at base, spine and first ray white; caudal golden, first ray 
white alK)ve and below; anal golden, the spines and first ray white; all the vertical fins narrowly c»dged 
with red; ventrals mostly pink, with golden wash on first rays; pectoral plain crimson; axil light rcnl. 

Color in alcohol, yellowish or orange white, the edges of the st^ales paler; some of the scales with 
small brownish dustings on the e<lges; edge of opi'Tcle black; opert^le and cheek somewhat silvery; fins 
all jMile yellowish, without dark edges. In some individuals the general color is more silvery, and in 
one example (No. 048()0) the axil of the pi»ctoral is somewhat dusky. In life the color is more scarlet 
than in M. murdjnn an<l the lins yellow, not red lu* in M. murdjan and all other Hawaiian spcM'ies. 

Af;fripnj<iv< rhri/sercH is relat^nl to M. murdjau^ from which it differs in the smaller wales, larger eye, 
less bla(;k in the axil, and the absence of blat^.k edgt»s to the dorsal an<l anal fins as in the life colors 
already note<l, the yellow lins iK'ing the most conspicuous character in life. It ri»aches a length of 9 or 
10 inches and apjH^ars to Ih> nHMlerately abundant at Honolulu and Ililo. 

Ty|)e, No. 50021), IJ. S. N. M. ( fiehl No. 03403), a s|KHinien S inches long, obtaini'd at Hilo, Hawaii. 
The numerous cotypes and the museums in which they have lM»en deposit^'d are indicated in the f(>l- 
lowing tabular list of s{>ecimens: 




Final dis[>o»<ition of siK-cinion. 









TviK*. No. ftUJ'iO. V. S. N. M. 





('otyi»o, No. 74r)2, 1.. S. .Ir. 
Univ. Mu.H. 










(:otyi.c, No. '22S.i. Am. Mus. 
Nat. Hi>*t. 






C^itviu', No. 2427:{, Ac. Nal. S<'i. 






Cotviw. No. 5(rMM). r. S. N. M. 





('^)lvpc, No. 394K, Field Col. 

IxMiKth. I^K-Hlity. Final disfHrsition of s{M><'imon. 

IH4'fu M. 

9 ' Hilo I Cotvpe. No. . M.(\Z. 

9 do I ('^>tyi»c. No. *Mf2, Ind. Tnlv. 

I Mus. 

s I Hom>lulu. <'otypc. No.'27(>2. T. S. F. ('. 

7 do (\»l"v|K', No. HIS9. Mus. 

s ' do.... CotyiK.', N«». 149().('al.Ac. S<-i. 

H j do 

7 I do .... 

H do 

S ■, 


15. Myripristis arg^yromus .Ionian i<: Kvermanii, new species. 

Hea<l 3.5 in length; depth 2.75; eye 2.4 in head; snout 5; uiaxillary l.H; mandible l.(»; interor- 
bital 3.75; I), x-i, 15; A. iv, 13; scales 4-33-5. 

Body rather hmg and compress(Ml, dorsal and ventnil outlines alxiut equally and evenly convex 
from snout to origins of anal and soft dorsal lins; head rather large but short; mouth moderate, max- 
illary reaching vertical at j>osterior edge of pupil, the cxpose<l portion broad, triangular, the upper 
edge concave, the end rouiuled and the anterior edge with short blunt teeth, strongest at angle; tip of 
iipj)er jaw with a shallow notch rouglH»ne<l at its outer e<lges; jaws (Hpial, lower fitting into the notch 
of upper and with 2 patches of strong blunt tooth-like tul)erclcs at it*< tip; eye large, its mid<lle above 
level of tip of upj>er jaw; interorbital spa<*e wi<le and slightly convex; 2 low, nearly parallel median 
ridges from tip of snout to nape, diverging slightly at their middle, another low ridge from above orbit 
ba(!kwanl to nape, and another backward around orbit; ridgis (m nap<> divergent; suborbital narrow, 
dentate on both edges; opercular Inuies all striate an<l dentate at the edges; opercle with a short, fiat, 
triangular spine; scales large, rough, striate near the edges whit!h are finely t<M»thed; a series of 4 or 
5 larger m()difie<l scaUw across najK', and a series of triangular sciil(»s along l)ases of dorsal and anal; 
alx>ut 10 scalers in front of dorsal; origin of «lorsid al>out over lower base of j)ertoral; dorsal spines 
slender, the first 3.2 in head, third an<l fourth longest, about equal to orbit; interval lK»tw«HMi dor-sjils 
very short.; anterior dorsal rays st)mew hat produced, their length e<jual tonnoutand eye; e«lge of fin 
concave, last rays nearly 3, or e(]ual to puj>il; anal sjunes graduated, the first very small, secon<l short 
but stout, thinl nnich longer and stoutest, its length 1.3 in eye, fourth still longt'r and more sh'uder; 
anterior anal rays produced, their length about equal to that ot longest <lorsal rays, free edge of fin 
concave; (^audal evenly forked, the lobes espial to length of head; {>ertoml long and pointe<l, reaching 
l>ey(md tips of ventrals, about 1.3 in hea<l; ventrals shorter, !.(> in hea*!, their tifw equally distant 
between their bases and that of first anal ray. 



Color in alcohol, pale yellowish-white, brightest above, more silvery on si<le and 1-X»lly; ojiereular 
bones with tine round brownish sfK-rks; edjre of oj)ereIe not black, scarcely <lnsky; axil dusky inside 
but not showing alM)ve tin; tins |nile yellowish-white without any <lark on e<lges. 

TyiH', No. 50031, U. S. N. M. (tiel<l No. 04821)), a tine si»eciuien 9.5 inches long, obtaine<l by us at 
Hilo, Hawaii. 

3/. nrgiirviHHn is relate<l to M. bcnnlti, but is <listingui8he<l by the more slender Ixxly, the al)eence 
of black on the opercle, and the i>aler axil. It does not apfx-ar to Ix? abundant and is repre»ente<i 
in our collections by only 8 siK.viniens. All the other si)eciiiiens are taken as tx)ty[H«. The data 
regarding each will be found in the following list: 






F'iiial db^iHusition of .specimen. 

Iitrht it. 


1 Ililo 

TviM', No. ."lUiWl, r. S. N. M. 



<'<»lvi»o, No. 7453. L. S..Ir.Univ. Mus. 

ti. .=» 

: i\n 

('«.lv|^. No. Uiyo, Bishop Mus. 



^^^tv|K^ No. 2-2^1, Am. Mas Nat. lUsi. 


! Honolulu.. 

CotvjK-. No. 2703, I'.S. F.C. 



Co^lH.'. No. . M. C. Z. 


! do 

CotvfK*. No. yKKi. Iiid. Univ. Muj*. 



Cotvpf, Nu.3lMy. Field V^A. Mu.-. 

16. Myripristis symmetricus Jonlan tk Kvennann, new sj)ecie8. 

Head IV2 in length; <lepth 2.4; eye 2.2 in hea<l; snout 5; interorbital 3.8; D. x-i, 15; A. iv, 14; 
P. I, 14; V. I, 7; scales 4-30-6. 

B<Kiy elongate, deep, coinpresseil, gn»atest <lepth alx)ut midway lietween origin of ventralsand anal; 
upi)er and lower profiles evenly convex; head comj)resse<l, as long as dee{), its width 1.7 in its length; 
snout short, broad, blunt, and steej>; upiHT profile of head straight from above nostril to occiput; eye 
very large, high, hardly impinging uikjii the upin-r protile of head, its diameter greater than poetocular 
region; mouth very large, oblicjue; mandible slightly projecting, and reaching i»osteriorly to l)elow 
posterior rim of jmpil; distal cxpandeii extremity of maxillary 2.,*^5 in eye; 8t»veral enlarged blunt 
teeth on outer front e<lges of mandible; te<*th in jaws, on vomer, and palatines very tine, in l>ands; 
tongue thick, iK)inteil, free; sulwjrbital rim narrow, finely .^*rrate; lower {K>st4'rior margin of maxillary 
smooth; Iij>s rather thick and tleshy; nostrils clo.vse together, jxisterior very large, close to front rim of 
orbit; l)ones of head all linely si^rrate; oj>orcle with wi'll-develo|fi*<j sj»ine; gill-oiH'iiing largt*, filaments 
large; gillrakers long, line, longest longer than longest gill-tilament; j>stm<lobran('hi:i' very large; dorsal 
spines slender, sharj», first 2.7o in head, si^'coikI 2.1, third 2, fourth \A\ tenth 6, and last 3.0; soft 
dorsal with anterior raVs elevated, pro<lnce<l inti> a point which projtM-ts l>eyond ti{» of posterior rays 
when fin is <lepressed, first my 1.4 in head, third l..*55, and last 3.7'">; anal sj)ines graduati^l to last, 
third enlarge«l, 2.5 in hea<l, fourth 2.i>; soft anal similar to soft «lorsiil, anterior rays j)n>duced, first 1.4, 
thini 1.3, and last 4.0; caudal elongate, deej>ly forked, the lol>es pointe<l, 1.2 in head, and reaching 
slightly Ijehind tiy>s of ventrals; ventrals sharp-pointed, 1.4 in head, spine 2.2; caudal (KHluncIe 
elongiite, compressetl, its length 1.8 and its dej>th 3.2; scales large, finely ctenoi<l, diH.'j) on middle of 
side; lateral line running t)bli<juely back, slightly cur\'eil at first, and jK>steriorly along upj)er side of 
caudal {>eduncle; 4 rudimentary, slender, shar|>-jH)inte<l, graduiiti*<l rays along upi>er and lower edges 
of caudal; Si*ales narrowly imbricate<l along middle of side. 

CoU)r in alcohol, i>ale straw-color; fins i^aler, except the anterior dorsal and anal rays, which are 
grayish; margin of oiKjrcle alxjve blackish; axil of jn^ctoral black. 

This specit^s was found lx)th at Honolulu and Hilo, but does not apfK-ar to lie abundant. Only 4 
specimens are in our collections: 






Final di-iMisiticn of sin't-iinen. 

5. 5 Hilo TyiH', No. riO»»J, r. S. N. M. 

5.5 do CotviK". No. 74.54. L. S. Jr. I'niv. Mu.**. 

5.5 <lo f'otviK^'. No. 2704, ['. S. F. ('. 

5.0 , Hom)Iulu OitviH.', No. 3y.'j0, Held Col. Muv. 


17. Flammeo scythrops Jonlan <k Kvermann, new specii*?. 

Head (measun^l Ui end of flaj)) 2.75 in length; depth 3; eye .Sin hea<l; .snout 4; maxillary 2.1; 
mandible 1.8; interorbital 5; D. xi, 13; A. iv, 9; tk'ales 5-4t^7, 5 rows on cheek; Br. 7. 

Body oblong, rather Hlender; dorsal outline gently and rather evenly rurve<l from tip oi snout to 
origin of Ht)ft dorsal, more nearly straight from tip of snout to na{>e; ventral outline U^kh eonvex; head 
long; snout long and pointe<l; maxillary broad, with a strong supplemental l>one whost* lower e<lge 
forms a broad angle; end of maxillary slightly concave; lower jaw long, much projecting, tip promi- 
nent; mouth large, not greatly oblique; maxillary nearly reaching vertical at i>osterior line of pupil; 
lips broad, rounded, and soft; eye large, lower edge of i)upil on axis of Ixxiy; interorbital si>m'e with a 
broatl, shallow groove between low ridges, one on each side; space Ix^tween ridge and eye with short, 
curve<l ridges; nape on each side with a grouj) of 8 or 10 short, sharp ridges, diverging bat'kwanl and 
ending in short, sharp spines; posterior part of supraocular with a patch of short spines; suborbital 
dentate on it« lower e<lge; prtMDrbital with 2 blunt prominences in front, a strong, reiMirvinl spine 
lx?low, ridges and spines on its upper surface; opercular lx)nes all strongly striate, the stria* ending 
in short spines; entire surface of interoi)ercle striate; operde with 2 strong sjnnes, the lower the 
stronger, its length 1.6 in orbit; preoperde with^ very strong spine at angle, it« length nearly 
e<iualing diameter of orbit, its surface striate, and its base with a series of small spines; under surfiu*e 
of dentary somewhat roughened; surface of articular In^ne much rougher; jaws each with u bnwd 
band of villifonn teeth, the outer series on upjKT jaw stronger; a narrow series on each palatine and 
a patch on vomer; scales moderate, the surfaces usually nearly smooth, the edges linely ttK)tlKHl; a 
series of strongly striate scales across nape, and a strong, striateii plate at shoulder; lateral line well 
developed, little arched, with alxmt 45 i)ore8; bases of soft dorsal and anal each with serii»8 of 
modified triangular scales; caudal with small scales on base an<l line sc-ales on membranes, extending 
well toward tips of outer rays; origin of spinous dorsal in aiivance of base of ixt^toral or over middle 
of upijer opercular spine; dorsal spines in a broad, <leep groove, UKKlerately strong, middle one 
longest, 2.3 in head, first a little shorter than snout, tenth more than half eye; dorsal rays longer 
than spines, longest 2.2 in head; first anal spine very short, second about 3 tiiiK»s as U)ng; third 
anal spine very long and strong, but little curvtHl, reaching past lyase of anal, its length 1.5 in head; 
fourth anal spine shorter and mon^^ slender, its length 2.25 in head and etjualing longest anal rays; last 
anal ray nmch shorter, IM in eye; pectoral long and slender, 1.2 in head, the tij> nearly reaching vent; 
ventrals shorter, etjual to snout and eye; (»udal forked, the lol)e« eipial, not strongly divergent, their 
length about e<iualing that of third anal spine; rudimentary caudal spines 5 alK>ve, 4 Ixlow, strong 
and sharp. 

Color in life, hea<l red above, paler on sides, nearly white IhjIow; tips of jaws rich re<l; side of 
l)ody with about 10 or 12 narrow yellow strijK's sej)arated by red or rosy stripes of about siime width, 
thos<> l)elow paler and somewhat purj)lish; un<ler parts purplish or [>inkish white; the strifH^s U^gin- 
ning at edge of opt»rcle and ceasing at biise of caudal peduncle, which is rich re<l aU^ve, lK?coming 
paler on 8i<le and below; membranes Ixitween the first and third dorsal spines rich bloinl-rtMl, thosi* 
between other spines white at Iwise, each with distal i)ortion lemon-yellow in front and red U'liind, 
last 2 or 3 membranes with little or no yellow; dorsal spines pale rosy, nearly white; soft dorsal, anal, 
IHJctoral, and ventral with rays rosy, membranes pale; vi'ntral with a little yellow at base; anal spines 
somewhat dusky; caudal rich ])l(K)d-red, jyaler distal ly; eye nn], a narrow yellow ring around pupil. 

Another example (No. 0:^041 ), much fadeil, was bright red; stripes on side e<iually bright golden; 
fins red; edges of dorsal membranes pale; no markings evident on fins. 

Co\oT in life of another examj)le (No. 03451), side with 10 or 11 longitudinal golden or yellow 
bands; spinous dorsal more or less white; membranes Ix^twcn^'u first an<l third dorsal spines more or 
leas deep vermilion, except the upper marginal portion lH»hin<l second sj)ine, which is white; a re<l 
blotch along margin of membranes just U^fore each of the other dorsid si>iiies. 

Color when fresh of another s|KH*imen (No. 034tK)), violet-rose with 10 strijyes of bright g(>lden on 
side; dorsal re<l, mottled with golden, the first two spim^a deep red; soft dorsiU an<l other fins rather 
light red without e<lgings, and Si-arcely darker behind thinl anal spine; pertx)ral and ventnds pink; a 
red dash across cheek, spa<;e above and Indow whitish; temporal region d<H'{> red; iris rtsl. 

All thest^ colors fade in alcohol and the fish Inn'omes a jmle yellowish white, the longitudinal lines 
on side showing faintly as duller and brighter strijH^s of yellowish white; fins all whitish or yellowish 
white, membranes of spinous dorsal whiter. 



The above description from the type, No. 50633, U.S.N. M. (field No. 03488), a specimen 9 inches 
long obtained by il** at Honolulu. An examination of our laqore j«eries of n>types shows but slight vari- 
ations, the characters appearing quite stable. In some examples the upper oi>ercular spine \» the 
larger, in others the 2 are equal; in 2 examples we find 3 oi>en"ular spines. 

This species ha.s been several times calle<l Ihtloceutrum nrgenlmm. The species des<*ribed under 
that name by Quoy *?t (Taimanl irom. New (iuinea reseml)les this in the slender Ixxiy and general 
coloration, but differs in having the lower jaw included, eye much smaller, mouth smaller, and the 
preopercular spine weaker. It was pmbably intended for some species with the lower jaw include<l. 

JlolocnUrns Here Lesson, from Tahiti, is more likely to l>e the prt»sent fish. It is figured as elon- 
gate, with the spinous dorsal low and the opercular spines equal. The plate is, however, too rough to 
permit certain identification and approa<-hes almost as closely to HoUM-ciUruJi diploj-iphuj* as to Flammeo. 

This is one of the most abundant spiH^ies in the markets at Honolulu and Hilo. It reaches a 
length of 8 to 10 inches. Our cx)llections contain the following specimens, a!l of which, except the 
first, are taken as cotypes: 


Length. Locality 






9. -25 


7. -25 







do — 



do — 



t • • ■ • V4" r • • • ■ 



do — 

FinAl dl*.pos.iiit)n of speii- 




Final disposition of spei'i- 

Type. No. ottJSi. V. S. N. M. 




Cotype. No. 7456. L. S. Jr. 


Univ. Mug. 

Coiype. N<». — . M. ('. Z. 




C<»t>-pe. NO.2705.U..S. F.C. 

Cotype. No. 74-'>5. L. S. Jr. 




Cotype. No. •2705.U.S. F.C. 

Univ. Muii. 


8. -25 


Colvpi', No. 24214, Ac. 


Nat. St±. PhilA. 



7. .50 


Cotype. No. 2285. Am. 

Col\-pe. No. 1491. CjiI. -Vf..<<i. 

Mus. Nat. Hist. 





Cotype. No. 7456, L. S. Jr. 

Cotyjie.No.SSirr. Field Col.Mu'^. 

Univ. Mus. 

Coty^K.'. No. 71.>;. L. .>. Jr. 
I'niv. Mu*. 









Cntvpe, No. 9>OI. Ind. Univ. 









Do. ; 






Cotype. No. 1«»1, Bu^hop Ma*;. 


5. W 








Cotype. No. 50<»4. U. S. N. M. 

Oiy^H 9. 00 
LXX. 1900. 492 . 


Honolulu ar 


Lchiier. l)enk>. .\k. Wi-v*. Wien, 

id I^y!<ant: not of Cuvier Ac 


18. Holocentrus xantherythms Jonlan ^^: Eveniiann, new sj)ecies. 

Head 2.S in length; depth .S: eye 3 in head: snout 4; maxillary '2.7; interorbital 5; D xi-14; 
A. IV, 10; scales 4-47-8. 

Bo<ly elongate, compressed, greatest depth alx)ut l)ase of ventral; upper pn>file steep; lower 
profile nearly horizontal; head compresse<l, its depth alxuit 1.2 in length, width 2.25; eye lar^, high, 
impinging ui>on upper profile in front, anterior, and a little less than fM>stocuIar region; snout short, 
pointed, its upper profile obliquely straight; jaws rather large, sul)e<:jual; maxillar\' reaching l)eyond 
front margin of pupil or to first third of eye, its dii«tal exj^nded extremity 2.7 in eye; supplemental 
maxillar}' largt*; lips rather thick, fleshy; teeth small, short, in rather broad l>ands in jaws and on 
vomer and palatinet^; tongue elongate, pointeil, free in front; nostrils close together, i)osterior, a deep 
cavity in front of middle of eye; interorbital space brr»ad, very slightly concave; preorbital with a 
large spine in front, its margins serrate; suborbital nam)W, with finely serrate margin; preopercle 
with a large dagger-like spine at lower angle; oi>ercIe with 2 similar spines cm upper margin, upper 
one much the larger; lx)nes of head with serrate margins; gill-opt-ning rather large, filaments and 
peeudobranchijc well develope<i; gillrakers short, comprt*sse<l, few, and much shorter tlian longest 
filaments; fleshy axillary* flap small; dorsal spines shari>-jK;)inte<l, first 3.2 in head, sei'ond 2.8, third 
1.9, last 7; anterior dorsal rays high, second 2.4 in head, thinl 2.2, last 6.5; thinl anal spine ver>' 
large, not reaching beyond sr»ft rays, 1.7 in hea<l, fourth 2.25; anterior anal rays longes^t, first 1.75 in 
head, seix^nd 1.9, last 6; caudal rather small, dt*eply forktni; j)ei'toral small, l.(> in head; ventral 
sharp-pointed, 1.4, spine 2; caudal pe<luncle elongate, i*ompre<s<eil, its length 2.1 in head, depth 4; 
scales rather large, ctenoid; lateral line nearly straight, running obli^juely down along upper side of 
caudal peduncle. 



Color ill life (No. 029S9), t>ri};ht iv<|, l)t»lly iimri* or k»SM Hilvory; alxnit 10 narrow Inii^itmlinal 
silvory HtrijH»H; upptTina**! pinkish; Hi«U» of heiui nilwry with piukiwh shadtv; a white ntrijK^ fnmi 
preorhital to Uuh» of pri'oi)en*ular spino; npinous dofHiil ilet'p hmI witlioiit streakn or hhuk inarkinj:, a 
white sjMit Itehind tin<t and wvtimi spines at Imtje, tii>s of thini to seventli sjiines whitish; soft dorsal, 
anal, i-amlal, and iM.vtoml plain ])ink; anal with nienihrane of third spine and first H)ft ray deep red; 
ventral pink, spine an<l first soft ray white, siTond soft ray det^p nnl anteriorly, iM»steriorly w hitish. 

Another example (No. (KiUM ), was row* nnl when fresh, with aUmt 10 very faint Ii>i:ht rosy streaks 
alon^ n)W8 of .»*eak»s, tlK»se nineh k»ss di.'ttinet than in other siKHties; elwek rosy with one hnKid 
obliiiue white tmnd; dorsal plain rhI, the tnend>rani« fading to white, no light striiies on dark aresu<; 
other fins ]>lain light red; membrane of fourth anal spine not darker; iris ))ink. 

Another exani)>Ie (field No. (KMliT), di»ep i-rimson when fn»sh, with 10 narn)w, shari>Iy defined, 
white strijH»8 along rows of soaU»s; an oblique white strijH* Ih'Iow eye fn)m snout to baw* of preojKireular 
8pine; dorsiil clear <Uh.»p re<l, (*loude<l with tlarker; st>ft dorsal, caudal, and anal light bright rt^l; 
mend>rane l)etwivn thini an<i fourth anal spines blootl-nul; jKH'toral dwp nil; veutrals rt^l, s]>mes 
white, their niembnuu^ bltMHl rinl. 

Color in aln»hol, pale bn)wn or bn)wnish while, waslu*<l more or less with silvery or brassy 
white; side w ith 9 or 10 longitudinal white strijK»s; fins pale. 

This siHM'ies h* n*latetlto IhtUHriUnmcmifir^ differing mainly in tlu» pnwMUf of two well-* levelojKHl 
spines on the up{KT margin of the oiM>rele. It is one of the most abundant of the family in Hawaiian 
waters. It is represi'nteii in our eolkrtions by 40 examples, jis follows: 

Kiuul diK{H).siUon i>f nikm*!- ' Fivhl 
iiu'ii. No. 


UUtil I 

(i , Honolulu. Tvi»0. No. THKir). r. S. N. M. ' (MJ-JT 

»i. n I dt» Cot VIM', Nt». .'HNiiu;, r. S. N. M. ' 

.>.:> I iio cotViH.'. N«».27(n.. r. s. F. r. i oi«,»mi 

!.:> ilo.... t'otviK'. No. -.VJol. Ficlil C«»I. OU>:is 

>fu^. (HOoy 

.\lV» «lo i'otv|K\ No. 7I.'»7, L.S.Jr. I'niv. OlVh»0 

, Mu-i. I iU\H\\ 

r..ri «lo , l>o. ,1 IM'.MO 

,'». 7:> «lo ColvjK'. N(». Iti'.rj. Hishi»|. M \is. OUW) 

.•..7:> ilo.... <'otvjK'. No. . M.r. /. ' OIJ.W 

."» do C«ilviH', Ni>. JJv;. Am. Mus. ! iM'l'M 

lA'iiKth. l^K'iiIily. 


0. 5 

Nnt. II iM. 



.\7.> do.... CotviK'.No.jrjl \ .\r. Nat.Sri. j Oiy7K 

I IfcVJd') 

(i. > 

I'hila. .... CoiyiM-. N«». •.•7;r.. r. s. F. c. 

• I. •! 

.'). ."> 

r **T 

it, it 

x 7:> 

Final «ll*«|Ntsiti«iii uf .K|HM'i- 

Honolulu. ('oiv|H'.Nu.ifcs<i\ Intl. Tniv. 

do , ('oiyiK',N«>. ir.rj.r«I..Vr.S'i. 

W.lh [ di» 

<i. 2.'> <lo 


t • • • • * ft' ' • ■ • • 

Kuilua ... 


KaiUm ... 


•t.l'y I liointlulu. 
."). r» <lo . . .. 


.Vnd ir> other o.xauipii's from llunolulu. Rin^'inK in K>nis'ili from 1 U) <!..'» inches. 

19. Holocontrus ensifer Jordan iV: ICvermann, new sfteeii^s. 

lleiid o in length; depth *J.7; eye ;► in head; Hin>ut ;».r»; maxillary lMi5; inten>rl>ital '»: 1>. xi, 15; 
A. iv, 11; r. I, 14; V. i, S; scales 4-17-8. 

Binly elongate, compn»ss4Ml, gri'atest depth at ventml fin; uj^ht profile <ltHide<lly more convex 
than lower; head compri'ssed, much k)nger than deep, |M)inted, its width a litth' more tlian half ita 
length; eye nuMlenite, alMUit 1.2 in i»ost»H*ular part of head, an«l ^lightly impinging u|M»n upjK'r profile; 
snout pointinl; mouth imxlerale, oblitpie; maxillary broad, with large supplemental Imuic di.<tally, 
t»<|ual to half diameter of eye; lips tliiek, fleshy; teeth minute, in broad bands in jaws, and <»n vomer 
ami i»alatint»s; tongue pointiil, free in front; n(»stri Is close together, iM>sterH>ra large cavity with s^^v- 
end small spiniv pngi't-ting over; pn*orbital with 2 large strong spines aii<l alnjul <> strong^'i rations on 
its margin; sulK»rbital rim narrt>w; bones of hea«l all mon» or less linely serrate, the ojH'rcle alH>veand 
pnHH»ercle K'low i»ach with a long, strong, dagger-like spine; interor)»ital spate broad, very slightly 
concave; a fleshy axillary flap; gill-oiK'uing large, filaments nio(U«nitely long, much longer than gill- 
rakers which an» eompn»sseil and not very numenuis; ps4'U<lol»nin<'hiji' large; spinous dors;il long, 
UKMubnuie lK'twiH»n spines not much incisiMl, fii>t 2.2, setond 2.1, thini 2, Uinl 4.2; anterior do rssd rays 
longi-st, fourth l.S, last 7.r>; thini anal spine largt'sl, 1.7''», fourth 2.:{; soft anal similar to soft tlorsal, 
thini spine not niiching Ivyond rays; caudal nitluT small, forked; |N'ctoraI !.:>; ventral 1.4, spine 2; 
caudal |K»<luncle i'ompn*:*H'<l, its length 2.2, tleptli 4; srak»s nither large, <tenoid: lateral nne an-lieil 
a little at first ami running down oblu|uely on upper side of taudal ikmIuiicIc. 



Color in life, bright re<i; side with atx>ut S yellow longitudinal liand^: spinour? <lonfal vennilion 
tingeil with yellow: soft dorval n»?*y with fnmt uiaivin white and l»eliind thisaUiw, re<l: anal whitish 
with red lK*tween thini spine and tirst ray: eaudal re<l, niargine*! aUive, and along the einargination 
with whitish; ptx-toral whitish with ri**! lines: ventnil n>sy with front margin white. 

Another example ttieM No. t>34o4K in life had yellow ami rwl lon^ntudinal liands alK>ve and yel- 
low and white below: spinoik' dorsal vermilion, other tins ixfl with white Ininlers. Another (field 
No. 0:^472 », was brilliant s<iirlet reil with 11 golden slreiiks alontr rows of Si-ales, upi>er 4 broadest, and 
third and fourth ni<")sl distinct and obli^jue: a white or goMen streak am »ss ehet^k ; tins plain searlet 
without dark i>atehes. CoU»r, when fresh, of anotluTsiHH-imen ( tieM No. (UU^*, bright re<l verging to 
scarlet; side re^l, with 4 golden stri|»es along luiok and ^ silver striiK*s l»elow these, golden and silver, 
ver}* bright; head crimson: a white l>and on cheek: spinous dorsal det*p sc-arlet with crims<»n etlge; 
soft dorsal light crimson with a white, then a dark crims«»n eilgi*: cau<lal bl<HKl re<l. etlge<i alxive and 
below with white, posterior part of tin abruptly fiale: anal with i»ale spim^, then bh»od re<l, then 
pinkish; ventral with white spine, then dark re<l, then pink: i:»ei'toml light re<l, axil deep reii. 

Color in alcohol, jiale brown or bn.)wnish white, the longitudinal liands on sides, together with 
scales on cheeks and oj>ercle, silvery: tins |»ale. 

This sjtecies was obtaineil by ib« at Honolulu and Kailua, and api>e:irs to Itecoiiinion at the former 
place. The collections otntain the following s|H*cimens: 






I 04960 


l>enKth., Lm-ality. Fiiml ili-|»iisitiim of >|nfim«*n. 


(i ' Honolulu. TypK*. No. .TOi37. r. S. N. M. 
6 KailuH ... Coiyjie. Nt>. 74•^\ L. S. Jr. 

I'niv. Mu*i. 
K.T5 Honolulu, retype. No.-TU^i. l'. S. K.C. 

S.5 I ilo 0»lvf»e. No. ItftKJ. Bi>hoii Mu>. 

S ! do TotyiK'. No. . M. «'. Z. • 

S.T5 j do , CmIvik.'. No. i>7. Am. Mus. 

Nat. Hist. 


^^y]'^ UuiiXh. l>HMlity. 


Finnl dis|Mi!*ition <if >peei- 

Inrh* n. 


Cotype. No. 24ilfi. \v. 

Nat. .<ci. I*hil«. 
y.7.'» do Cotypt*. No. yNJ6. Ind. 

v^.5 do C'ot.viK*. No. 3*»2, Field 


Family CAK.WGID.-E. The Pampanos. 

20. CarangxLs elacate .Ionian iV Evennann, new sj»ecies. 

Head o.6 in length; ilepth 8.4: eye 4.5 in hca<l: .^nout ."^.S: intcrorbital .S.Siu snout: maxillary 2.1; 
preorbital S.5: mandible !.!♦: 1>. vii-i. 19: .\. ii-i, IH; S4Uti'S l*s. 

Bo^ly slender, compressiil. imt gn»atly clevatc<l: snout rather short, protile asiendinir to na|« in a 
gentle curve, slightly trenrhant: m<juth larirc, >lightly oblique: lower jaw «<imewhat projecting; max- 
illary reaching iKi.'^terior ctige of orbit, its width at tip l..> in orbit: supplemental maxillary' well 
developer!, its width .S.2o in entire width: gai>e reaching vertical of iK>sterior e«lire of pupil; villifonn 
teeth on vomer, palatines and tongue, tln»>e on jaws in a single n.>w, small and somewhat c-auine-like; 
eye largi', anterior; adijH>st» eyelid stn>ngly develoi)e<l l»ehin<i; supraoi'ular region with two ridges, 
extending to humeral region, the lower the stronger; jx>sterior lialf of Ixnly, lieginning at origin of 
soft donal, long anil gently taj>ering to caudal peiluncle: caudal iKNhuicle nmch depn^ssetl, its least 
depth Si-ari-ely half its least width; distan«-e from liase of last dorsal ray to origin (»f caudal tin equal 
to snout and pupil: tins small; r»rigin of .spinous <lor8aI jK^sterior to base of |)ectoral by a distam-e 
equal to eye; longt^st donal spine slightly grt»ater than snout: anterior rays of soft dorsiil somewhat 
produtt^l, alxiut 1.8 in heati; anal similar to s<.>ft <lorsal, ib* origin under eighth soft dorssil ray, anterior 
ray pn>luceil, but s^-an-ely equaling longest soft doreal rays; caudal widely f«>rke<l, lobes ap}iarently 
equal; pectoral long and falcate, reachimr juL^t origin of anal. exaH*<ling head in length by 0.65 diameter 
of eye; ventrals short. *2.4 in heail: scales rather large, a low . •sheath at l>a.«HM>f si>ft dorsal and anal 
anteriorly; breast entirely s<-ali^l; lateral line stnmgly art-heii al.H»ve i»ectoral, joining straight j>ortion 
under sixth dorsal ray, chonl of an.'lie«l i>ortion l.»> in straight |>art. 

Color in alcohol, nL-^ty olivac^nnis alnn-e, jvaler on side lx4ow lateral line, l»elly white; top of heail 
dark olive, side an<l lower jaw lighter, with stning brassy tinge on pK«it*K-ular and on lower }>ortioiis 
of oi»ercle: lower jaw pn^fusely covere<l with tine brown jM.iints; a Idack s|Hit at upjuTend of »»peri*ular 
0})ening; axil black; vertical tins all more or less dark: pnMluce«l i»art of S4.»ft dorsal almost black, low 

F. r. B. iyo2— LJ 



part of soft <lor8al black at Imise, then ligliter, narrowly ti}>]»e<l with dark; anal <lark brown, with a 
Bubtenninal strijx' of yellowish white alonjr wlj?e of fin; j>ei'toral and ventralH pale. 

The al>ove d(*scription bai*e«l njMm the tyi>e, No. 50638, T. S. N. M. (tield No. 04452), a larpe 
exami)le 27 inches long, from Honolulu. 

This speities somewhat resembles Carmi(piJi nianjinnfHify from which it differs in the nmch more 
slender Ixxiy, larger eye, and <lark anal fin. The ty{H» is the only examjde obtaincnl 

Family SEKRANII).4v. The Sea-basses. 
21. Pikea aurora Jonlan & Kvemiann, new sjHM'ies. 

Head 2.5 in lenjrth; depth 3; eye 5 in head; snout 4; interorbital <>.2; maxillary 2.25; I), viii, 
13; A. Ill, 8; scales 5-55-22; Br. 7; jrillrakers short an<l rather weak, alnrnt 9 • 5. 

Bo<iy moderately stout, the back slightly elevated, hi'ad rather long and p(»nte<l; snout depressed, 
the anterior pn>tile nearly straight from tip of snout to (H'<'iput; mouth large, maxillary rea<'hing 
posterior margin of puj)il, supplemental lK>ne not develoi>ed, the tip bn>a<l, 1.5 in orbit; mouth some- 
what oblique, the lower jaw strrmgly projecting; t(H»th in broiKl villiform l^ands on jaws, vomer, and 
palatines; tongue naketl; eye mcnlerate, high up, chiefly al)ove axis of IkmIv; anterior nostril in a short 
tul)e at cnlge of prenasal; [)Osterior nostril small, nmnd, near upiKT anterior inlge of orbit; (Mlgi» of 
preo[)ercle slightly dentate, espwially on lower arm; oiR^rde ending in a bntfwl flap with a weak, flat 
spine; peeudobranchia* rather small; int^irorbital low, very little convex; caudal peiluncle .''tout, com- 
pre8»e<l, and very deep, the depth e(]ualing snout and eye; fins rather small ; origin of dorsal posti*ri«)r to 
that of i)ectoral, slightly nearer Iwise of last ray than tip of snout; dorsal spines low and weak, the thini 
longest, 3.6 in head; soft [portion of <lorsal sr^newhat elevate<l an«l ixnntAHl, with longest niy 1.9 in he^l; 
anal similar to soft dorsal but smaller and somewhat iM)sterior, fifth ray 2 in heiul; caudal truncate or 
slightly lunate; ventrals short, not nearly reac*hing vent, their length 1.75 in head; iK^rtcjral rather long 
and slender, reaching origin of anal, its length al^nit 1.4 in hea<i; scales rather small, finely ciliat4.», 
somewhat loose; entire head, excej)t interorbital, snout, and under parts, s<-aled; latenil line well devel- 
oped, comi)lete, with a stnmg arch alx3ve the pectoral and distinctly decurvtMl under last dorsal ray. 

Color in life (field No. 0:^:^42), top of hea*l, upi)er half of anterior jiart of IxmIv, and whole |)osterior 
half of bo<ly j)ale rosy; lower j)art of head, and lower parts of ant(?rior half of lH)dy white with faint 
rosy wash; top of hea<l and luwrk in front of dorsal verniiculat^Ml with grt»enish yellow lines; middle 
jHjrtion of uj)per jaw yellow with a broad sulphur-yellow strii>e from it to eye, then back of eye to 
opennilar opening; a narrow sulphur strij^e on j)osterior e<ige of maxillary and i-ontinui^l interruptiHlly 
downwanl and l)ackwanl across cheek to opt^rcle; a few small yellow sjiotw a<Toss cheek lx*twiHjn 
the two stripes; ti[) of lower jaw yellow; yellow of Iwick in alKmt (> indefinite lines; dorsal pale rosy, 
spinous part greenish yellow at base, this extending toward tip jMisteriorly and fonuing a submarginal 
yellow stripe cm soft part, narrowly lx)rdenMl alM)ve by rosy; n»st of fin rosy; caudal dark rosy, paler 
tf)ward tip, then with bla(;kish red inlge, a grtH»nish yellow striin.' along upiK»r and lower margins nar- 
n)wly tHlge<l with rfw?y; anal yellow anteriorly, n'st of fin jwde rosy; {>ec^toral and ventrals pale n>sy; 
yellow of lower jaw Inmnded by rosy, rest of jaw ami chin whitish; some examples with iM>sterior 
half of side with s<'attere<l small greenish yellow sjmts, these extending on caudal; eye with a broad 
brown bar through the micMle, white above and l^elow. 

Color in alcohol, jiale yellowish white, lighter 1k»1ow; Inxly, esiH'cially jxKsteriorly, caudal, and soft 
dorsal fins with numerous small distinct brown spots; hetul pale, a white line extending along up{H^r 
edge of maxillary and acToss cheek to ojwrcular o{)i^ning, a similar but less di.^tinct white Hih' from t»ye 
to upper edge of gill-o[)ening; between these 2 a few white sjK'cks; all the fins except cauiial and soft 
dorsal plain yellowish white. 

Four specimens of this interesting and hamlsome sjuH'iiw were obtaintHl by us, 2 at Honolulu and 
2 at Hilo. Four others are in the collection nuuU? at Honolulu in 1898 bv Dr. \Vo(m1. 









5. 75 




Hilo ' 


Fiiiiil iliMfMisition of HiH'ciiiu'U. 

TviK', No. 5()«i75. r. S. N. M. 
('^>ty|.<». No. :W7I, Flcl<l Col. Mu.s. 
Cotypo, No. 7is^l. L. S. Jr. riiiv, Muh. 
Ctitype, No. 27:M, T. S. F. C. 




o. I'..I. 

Inrht ;*. 














22. Anthiaa kello^^ Jordan «fe Everiuann, new sjHH'ies. 

Hea<l 2.5 in length; depth 2.5; eye 4.5 in head; snout 3.6; maxillary 2; interorhital 5.4; I>. xi, 15; 
A. Ill, 7; P. 15; sc-ales 4-36-10; gillrakers IH -4. 

Body short, <ieep, and compres!«e<l; dorsal rmtline greatly art'he<l, j>rutile from orij?in of spinous 
dorsal to tip of snout nearly straight, l>eing *rently concave over inten)rl)ilal sj^ace; ventral outline 
nearly straijjht; caudal {)e<luncle c<»nipressiHl, its jrreatest depth '.\ in hea«l; head lonjrer than deep; 
snout bluntly jM^inte*!, lower jaw prominent, slightly the longer: mouth large, nearly horizontal; a 
narrow band of small, shar]), conic teeth on palatines, a small pat<-h on vomer, a band of canliform 
teeth on upper jaw, a narrower l>and in lower jaw; several largi* canine teeth in each jaw anteriorly, 
3 of these close together on mi<ldle of each side of lower jaw, tlu*se ho<»ke<l backward; 6 or 8 large 
pores on lower side of mandible and several on upper j>art of snout; maxillary reat-hing to j>osterior 
edge of orbit, its greatest wi<lth 1.5 in eye; e<lgt» of preoj»en'le alnive angle and t^ige of opercle below* 
the upper middle of Iwse of {XM'toral <lenticulate; 2 bn^ad of>ercular spines, the uj>per the larger; eye 
anterior, its lower edge on line with uppi»r l>ase of i)ectoral; tins large, the stvond soft dorsal ray and 
upper rays of upj)er cau<lal lobe l)eing pro<lu(vd ea^'li as a filament, the dorsal filament l^eing produce<l 
half its length l)eyond rest of fin; dorsal .^j>ines stout an<l strong, the first spine 2..S in thircl, the fifth 
l^eing the highest, 2.5 in head; base of spinous dorsal 1.15 in hea<l; bai*e of soft dorsal 2.3 in head, its 
fourth ray 3.5 in head, the last ray 1.4 in fourth; cau<lal truncate, the lower rays ppHlucetl slightly as 
a filament, but not nearly s<_> long as the upi)er lol)e; Si*cond anal spine longest, 2.5 in head; second 
soft ray longest, 2 in head; i>i»ctoral very long and large, reaching to origin of soft anal, the eiirhth 
and ninth rays from the top the longest, 1.4 in head; s«-al(»s large, finely ctenoid, in regular series; 
entire Ixxly and hea<l scaled; l>asal iKjrtion of all fins except spinous dorsal with small scales; lateral 
line strongly convex, not concurrent with the dorsal j»rofile, I becoming straight on middle of caudal 
peduncle; one row of scales Ix'hind tip of last dorsal ray. 

Color in alc<:»hol, f»ale brown, the tins lighter; in life, red. 

Only three si)ecimens of this s|x?cies were obtaine<l, all having Ix^en taken with the hook in deep 
water off Kailua, in si»uthwestern Hawaii. It is allie<l to AiUhiaj< jnjp^niicus Steindachner <fe Doderlein. 

Nametl for Dr. Vernon layman Kellogg, profcss<ir of entomology in Stanford University. 

Sjj^' Length. I>N-ality. Final di.-fKtsition of >pe<iineii. 

av.>7*< 7. :.'> offKailim TvfK-. N... .Vi»V4J. l'. S. X. M. 

CCiTWi s 'U* rotvi«'. No. lUAK L. S. Jr. Univ. Mns. 

(iV277 S..S iio roiviK. No. jTii. r. .<. F. r. 

Family APOGOMD.^. The Kint; of the Mullets. 
23. Apog'onichthys waikiki Jonlan ik Kvennann, new species. 

Heail 2.4 in length; depth 3; eye 3.2 in hea<l; snout 4.6; interorbital 6; maxillary 2; I), vii-i, 8; 
A. II, 7; scales 2-24-5. 

Botly short, stout, and compresse<l; <lorsal outline strongly arche<l from tip of snout to jM^sterior 
base of soft dorsal; ventral outline comparatively straight from tip of man<lible to origin of anal; vent 
immediately in front of origin of anal; caudal j^e^luncle deep and compresscMl; head rather large; 
mouth large, slightly oblique, jawseijual, maxillary reaching posterior e<lge of pupil; eye rather small, 
slightly above axis of IxmIv; intenirbital spacv narrow, little convex; opercular and preorbital liones 
entire; a l>and of small villiform tt»eth in each jaw, and on vomer and palatines; fins mo<lerate, origin 
of spinous dorsal nearer ba.<»e of last soft ray than tip of snout; first donail spine very short, second 
about half length of thinl, which is equal to eye an<l snout; l»ase of soft dorsal e<iual to depth of 
caudal i^eduncle; longest <lorsal rays 2.25 in hea<l; cau<lal rounde<l, its length 1.75 in head; origin of 
anal slightly iKtsterior to that of s^)ft dorsal, its longest rays 2.4 in heail; pectoral slender, reaching 
past origin of anal, its length 1.5 in head; ventrals short, Imrely reaching origin of anal, their length 
nearly 2 in head; .'H'ales large, weakly ctenoid, firm and somewhat deei>er than long; lateral line 
strongly develo|)e<l, following outline of liat-k until under last dorsal ray, where it curves downward, 
following middle line of caudal peduncle to biise of caudal fin. 


Color in alcohol, head and IkkIv rather dark brownish, a lighter eropsband around body at nape 
and arross opercles; another light ban<l surnmnding b<Kly between the 2 dorsal fint?; 8 dark brown 
lines radiating from the eye, the first downwanl aeross cheek to tip of maxillary, the Hecond 
backward acnws cheek towani l)a8e of jXH'toral, the thin! upward and l>ackwani to origin of lateral 
line; spinous dorsal blackinh, esjKH'ially on last spine; soft dorsal, anal, ami caudal dusky, narrowly 
e<iged with white; i)ectoral i>ale, crossed by al)out 6 ol>s<*uri^ brownish <!n)ssbars; ventrals blac^.k or 
very dark brown, the outer rays somewhat paler. 

The above description is based ujxm the tyi>e. No. 50H39, V. S. N. M. (field No. 20), a s[HH'imen 1.5 
inches long, obtained from the coral rocks in fn>nt of Waikiki, near Honolulu, August 22, 19()t. 

This species is distinctly related to A. alutuj< of the coast of Florida, fnim which it differs mark- 
edly in color and in the more slender body. Only one 8i>e<'imen was obtained. 

(Henni FOWLEKIA Jordan ft Eyermann, new genni. 

Fowleria Jonlan & Evermann, new geiuis of AjH)fjonida' (iwrita). 

This genus differs fnim Aj}Offoiiu'hthtjH only in the charai'ter of the lateral lin«', whi«rli is develoiHHi 
only on the anterior part of the Ixxly. 

Several species otunir in crevices of coral ro<'k in the South Seas. All of them art* of very small 
size and some are brightly colore<l. 

This genus is named for Mr. Henry W<H?d Fowler, of the Aca<it?my of Natural Scieni^es of 

24. Apo^n snyderi Jonlan & F.vermann, new s|HM*ies. 

Head 2.7 in length; depths. 1; eye*i.7 in heati; snout 8.7; inten>rbital4.5; maxillary 2.2; mandible 
2; gape 3; D. vii-i, 9; A. ii, S; C. 17; P. 10; scales 2-25-5; Br. a 

Body short and stout, moderately comj)rt»ssiMl, the <lorsal and ventral outlines al)out €H]ually 
curved; hea<l rather laiye, conic; snout conic, the anterior pn)file very slightly curveil from tip of 
snout to origin of spinous dorsal; mouth oblicpie, jaws sul)e<|ual, the lower slightly include<l; maxillary 
long, reaching not quite to posterior tnlge tt( i)Upil, its width at tip 2 in eye, supplemental bone well 
develoi>ed; interorbital sj»ace rather broad, slightly convex, preorbital narrow, least wi<lth 3 in eye; 
teeth on vomer and jaws, the latter in villiform bands; none on palatini*s; gillrakers slender, 10 on lower 

in neacl, secona z in uie lounn, st»veuiii -: in m.Toim; nn^i wiii. rays longesr, i.r* in nea<i; caudal deeplv 
einarginate, longest rays alK)ut 1.6 in head; anal similar to soft dorsal; somewhat smaller, its origin 
under last rays of soft <lorsal ; ventrals jMwnted, scarcely reaching vent, \A) in hea<l; iK^toral n^aching 
vertical of vent, 1.7 in head. 

Color in alcohol, pale yellowish brown, darkest alM)ve; a darker brownish band extending from 
upper edge of ojwrcle along siile, just above lateral line, to po.Hterior t^lge of soft dorsjil; another 
broader, more distinct brown band from tip of snout through eye and along miiMle of side to baMe of 
caudal fin, cx^vering lateral line on caudal jKnluncle; cjuidal p«Hluncle at bii.«t» of caudal fin with a 
))road duflky crossbar, usually darkest on upper half, *^)metim(»s obMcure, sonietimes with a darker 
blotch or spot in the upi)er portion; upi^T part« of hea«l covchmI with fine dark bn)wn punrtulations; 
lower jaw similar, but somewhat paler; membnuu^s of anterior 2 or 3 <lorsal Hpines black, others 
finely punctulate; soft dorsal pale at base, al^>ve which is a ])r()ad indistinct dark crossband, the 
color confined chiefly to the interorbital membranes, this color extvn<ling to near tii>of last rays; outer 
part of soft dorsal pale; anal similar to sr)ft dorsal, the black bar narrowiT and nejm»r base of fin, n»st 
of fin white; caudal dusky on membram^ of outer I or 2 niys, tlu^ fin otherwise white, with a few fine 
punctulations on the interradial membnint^s; ventrals pale; distal |»arts of the first and second niys 
and their connecting membrane black; i>tH"toral pale; axil and basi' of in^'tonil somewhat dusky. 

Color in life (field No. 11W, (). P. J.), pale red; 2 longitudinal pearly Uiu^ on hiMly; lirst dorsal 
with a dusky olivaceous anterior Iwrder; white lines along ft^urth, fifth, sixth, ami seventh spines, the 



membrane olivat^eouH; 9ect>n<l dorsal with many white and some olivaoeoiis npots*; anal with a dusky 
line along base, the distal |)art rt*<l; l» of caudal dupky, re-^t of fin pale re<l; ventral with a white 
spot near tip; pei-toral pink; iris yellow.. 

Another example (tiehl No. 03499) wa« copi)ery brown when fresh, with trace of dusky band along 
side; a faint blaok liar at liast* of caudal, forming a black spot above en<l of lateral line; some dusky on 
opercle; first dorsal dusky; second dorsal brownish re<l with some dark; anal same with a basal flesh- 
c<dored bar Inflow it; caudal re<l<lish brown; ventrals same, with first ray pinkish and dusky behind 
it; some dusky on opercle. 

This si:>ecie8 reaches a length of alxmt 6 inches. It was obtaincnl by Garrett in the Hawaiian, 
SrKriety, and Paumotu islands. C>ur collections contain numerous spiecimens fn>m Honolulu and Hilo. 
We have examineil also 12 sj^et-imens in the collection made by Dr. O. P. Jenkins. 

Thb? species cU>sely resenibles Ajn^jon menr}*emuJ!t, from which it <liffers chiefly in coloration; the 
black caudal crescent, which is sucli an excellent <listinguisliing mark in .!. mmesemuji, is wholly alieent 
in thb« species; moreover, the 2 silvery lateral l>ands, which become dark brown in spirits, are not 
found in .1. menejtemu:*; and the black on the anal an<l soft dorsal is less cf^nspicuous in A. tenyderi. It 
l>elongs to the subgenus IYiMki}K}fj(m of Klunzinger, having Iwth limlw of the preopercle serrate. 

This species is figured by Bleeker, Day. and Giinther, the figures of Bleeker and Gunther being 
colore<l. The.l)eHt figure is that of (Tunther in Fische der Su<lsee, who calls it AjHpgim freimUt*^ but the 
species originally thus name<l seems to Ije quite different, &* Bleeker has alrea<ly noticed. 

Named for Mr. John O. »Sny<ler, assistant {)rofessor of zoology in StanfonI University. 

Our collei'tion contains the following specimens of this si»ecit^: 





Final disp4>MJ 

tion t>f sfteoimon. 




Final di 


ition of 8pe<*imen. 






T\-p*». No. 50040. r. S. N. M. 



- Honolulu. 




Cotyi»e. No. 

S0»V41.r.S. N. M. 






do ... . 


459. L.S.J r.rniv. 











9>07. Mus. Ind. 








CotviK', Xo. 

•2709. r. S. F. C. 



. Hilo 



9S0H. Mu!«. Ind. 




CotyiK'. No. 

•2710. r. s. F. r. 





CotyiK". No, 

— . M.r. z. 



, do.... 



. :«')6. Field Col. 


4. 25 


Toivi^e. No. 

. M.C. Z. 






Cot^lK-. N«.. 
Nat. Hist. 

■2!^^>». .\n». Mn-. 







3957. Field Od. 




CotviK*. N(». 

'1>>\\ .Vin. Mu^. 

051 4M 

4. 75 




149:^. Cal. .\e.Sci. 

Nat. Hi.^t. 

(i'>l 19 





1494,ral.Ac. Sci. 




r4>tV|K^. No. 

•24217. \v. Nat. 


4. 75 




lem. Bishop Mus. 

si. rhila. 






1695, Bi.«hop Mum. 




('otyin' . No 
S<i. I'liila. 

'242 l.s. Xv. Nat. 

<>. I'.I. 














• > 










do 1 














4. 4 





•1 4J.S 



















3. 75 



:i. 5 


'' f.7:{ 





« Dr. WtMxi. 

' .Ionian A 


Apttgim /nwitmt. (iiinther. Fis<'he <1«t Siitlsee. ». 19. laf. 19. ti^'. A.l'*7:i (Hawaiian. .'^iXMety. and Paumotu islands): Stein- 
daehner, I>enk.»-. -\k. \Vi«w. Wieii. i..\x, l'.^(ii». 4M ( Honolulu), not Aift^f^nt /nnuUnn Valeiieieunes. Nouv. Ann. Mus. 
Hist. Nat. Is:r2. 57. pi. 4. tiif. 4. nor of Klunzinger. 

Family FKI.MANTHII).4:. The Calalufas. 

25. Priacanthus alalaua .Ionian & Evermann, new si>ecies. *M/<iAiii/i." 

Head 3.2 in length; depth 2.<)5; eye 2.4 in head; snout :\Jk maxillary 2; interorbital 3.8; I), x, 
14; A. Ill, 15; sc-ales i:^S5 to 90-45, 70 iM>res; Br. i'y; gillrakers, alxuit 22 on lower ami. 

IVnly short, deep, <*ompressed, ovate; up|»er pn)tile of head nearly straight; snout very blunt; 
mandible prominent, produced; mouth vi^ry oblitjue; teeth small, sharp, in bands on jaws, vomer and 



palatines; tongae rounded, free in front; maxil1ar\' reachinf? almost to front margin of pupil, its greatest 
width 2 in eye; edge of preoi>erole finely serrate, with a sharp, tiat, serrate*! Hi>ine <lirecte<l liackward 
at angle; margins of interojiercle, sulK>i)ercle, and opercle entire; ojHTcle with an ol>Hcure flat spine; 
interorbital simcv slightly i*onvex; eye very large, it>* lower wlge a little al)ove l>ascj of i)e<;toral and in 
line with axis r»f l)ody; nostrils small, close together, the anterior with elevatetl rim; posterior nostril 
oblong, with broad flap; gillrakers rather slender, alnnit 22 on longer ann of flrst arch, longest about 
3 in eye; origin of spinous dorsal over upjHKr ba**t» of iKH'toral; dorsal spines rather uniform, the 
longest about e<iual to orbit; soft portion of dorsal wniiowhat elevated, rounde<l, fourth ray 1.7 in 
hea<.l; anal spines rather stnmger than th«>se of dorsal, thin! the longest, 1.1 in orbit; soft jx^rtion of 
anal similar to that of soft dorsal, rays of al>out etiual length; caudal truncate, the middle rays slightly 
greater than orbit; i)ectoral short, bluntly iH)inte<l, not ri'aching tip of ventral, length 1.4 in head; 
ventrals longer, just reaching base of second anal spine, their length 1.2 in head; ventral spine alK»ut 
1.25 in longest ray, or 1.7 in htwl; g«iU»s small, firm an<l rugose, those* of lateral line somewhat 
enlarge<l; entire hea<l, as well as l>ody, densely wealed; lateral line rising abrui>tly for 6 or 7 iM>refl 
from gill-oi>ening, theniH? concurrent with back to caudal peduncle. 

Color in life, silvery, light olive alx>ve, somewhat flushwl with rtnl in irregidar blotches; chin rvd; 
spinous dorsal olive-yellowish, esi)ecially on eilge; ventrals bla(*k, rays whitish; fins unspotte<l. 

Young of 4 inches in length art» dirty gray, browner al)ove, with no trace of reil in life; some bn)wn 
spots along lateral line; fins dusky, anal and ventral darkest; iris a little brownish red. 

Color in alcohol, plain yellowish-white: spinous dorsal and anal somewhat dusky; ventral mem- 
branes black, the rays white, other fins pale yellowish-white. In some examples the color is nmch 
more flushed with rtnl, espc^cially al)ove; the re<l paler ami more evanes(»ent than in the other sjHH»ii*s; 
fins re<l, unspotted; the spinous dorsal eilged with golden; upiwr lip golden; ventral membrane 
black, jKJctoral pale. 

There set^ms to Ik* but little variation in this 8i>ecies; the younger indivi<1uals a]>]H*jir to Ih» more 
brightly colored or with more evident wash of nnl than was shown in the tyjH'. , 

We have examine<l the following sptnumens: 




Iiw.hftt. j 

M.'i'» Froiiolulii 

.Mo .. 

Final clis|N»sitioii of 8|K'i'inH'ii. 

Tyiu'. No. ftO(M3. W S. N M. 
<"oty[H«. No. 74«il, L. S. Jr. Iniv. Mum. 
(;otviM'. No. 271-2, r. S. K. C. 

Family LUTIAXID.^. The Snappers. 
Oenm B0WEB8IA .Ionian <% Evermann, new genus. 

Boirersia Jordan & FA'cmiann, new genus of Ltii'unihh' {violniwnift). 

Iknly l((ng, rather slendt-r and nirxlenitely comprt»ssed; top of head evenly roundt*<l, the supra- 
o<.*cipital cn»st exten<ling forwanl on cranium; jaws equal, lower not projet'ting; bands of villiform 
teeth on lK>th jaws, the outer s«»rifs S4imewhat enlarge<l and <>anine-Iike; villiform tt^th on vomer, 
]»alatines, and tongue; maxillary slipping for its entire length under the rather broail i)reorbital; eye 
laiye; ojiercle entire, ending in 2 tlat, ol>scure sjiines, the space l>etween them deeply emarginate, 
but filled by sr)ft membrane; pn^oiK-n-le scanrc*ly dentate; dorsal fin C(mtinnous, the hist ray i)r<MlurtHi 
nearly twice length oi i»rectHling one. 

This genus is R*late<l U\ ApRihin^ with which it agnH»s in the ]>resence of villiform teeth on the 
vomer and palatini*s, but fn)m which it differs in having well-d<*velo|H'd tivth on the t(»ngue, and in 
the productMl last dorsal an<l anal ray. Two sjxM'ies are known. 

We take much jtleasun^ in naming this new gt'iius for the Hon. (Jeorge M. Bowers, I'nitefl States 
Commissioner of Fish an<l Fisheries, in re<'ognition of his active and intelligent inten*st in i)roinoting 
scientific work, es}K»cially the investigation of the a4]uatic resources of the Hawaiian Islands. 

i\. St'Hles ratluT l«rjr<*. alMHit iW in lalonil line; |)rt'«>rhital hrtMni. 7.7.'» in hfu«l 
(la. Scales smaller, about ti-s in lateral line; |>ri>ort)ital narrow. 10 in IuncI 



26. Bowersia violescens Jonlan & Evermann, new ppeciea. ^^OjHikapaka.*' 

Head 3.25 in lenjrth; depth 3.5; eye 4.4 in head; snout 3; maxillary 2.6; mandible 2; inter- 
orbital 3; preorbital 7.7.'>; whales 8-<>0-15; D. x, 10; A. iii, 8; Br. 7; ^illrakers 5-rl4. 

Body long, rather slender, mo<lerately (•(>niprea*«e<l, tapering gradually into the rather long caudal 
pedunole; head large, longer than dc^p; snout moderate, rather bluntly conic; mouth large, maxillary 
reaching anterior third of jmpil, slipping for its entire length under the thin e<lge of the rather broad 
preorbital, the width of its tip 2 in eye; mandible strong, but not projecting; broad bands of villiform 
teeth on jaws, vomer, i>alatines, an<l tongue, the outer series in the jaws slightly enlarge<l and canine- 
like; eye large, its lower edge in line with axis (»f Uxly; interorbital broad, gently convex; anterior 
profile but slightly curve<l from tip of snout to najK?, thence more strongly arche<l to origin of dorsal, 
descending in a long, low curve to caudal i)e<luncle; ventral outline but slightly convex; caudal 
peduncle rather long, 2 in head, its least width al>out l.(> of its leiust <lepth, which is 1.8 in its length, 
measure<l from base of last dorsal ray to base of supporting caudal rays; gillrakers few, rather strong 
and short, the longest about 2.6 in eye; opercle smooth, ending in 2 flat, ol)s<'ure spines (more strongly 
develope<l in ea<!h of the cotyi>es); prtn^percle obs<'urely .serrate at the angle (more distinctly so in the 
cotyi)es); fins nnxlerately develojieil, the dorsal fin continuous, without not<-h, its origin over base of 
f)ectoral an<l ecjually distant from tip of snout and l>a."^ of fourth ray, length of entire Imse of fin and 
to tip of last ray twice length of h(»ad; first dorsal spine mo<lerately short, cla**ely bound to the second, 
whose length excee<ls it by al)out one-half; seventh dorsal spine longest, its length e<jual to that of 
snout; last dorsal ray pro<luce<l, its length alxmt 1.7 times that of the preceding; anal similar it) soft 
dorsal, its origin under base of third or fourth dorsal ray; first anal si)ine very short, thinl longest and 
strongest, its length equaling diameter of eye; la^tanal ray pro<luce<l, its length t^jualing that of produced 
dorsal ray; caudal rather widely forkeil, lol>es alK)Ut equal, their length, nu»asure<l from base of first sup- 
porting ray, equaling heml; ventrals jKnuttnl, their tips not reaching vent, length 1.4 in head; pt^ctoral 
long, slightly falcate, the tip alK)ut reaching tii)s of ventrals, its length a1)out 1.2 in head; 8<*ales lai^, 
deeper than long and rather loose; cheek and ojKfrrles scale<l, 5 rows on chtn^k; a large lx)ny humeral 
si'ale, from which extends to nape a s^'ri€*s of somewhat UKKHfied scales, in front of which is a patch 
of onlinary scales; lateral line ci»niplete and well develojjed, bt^inning at lower e<Ige of humeral scale 
and following curvature of l>ack to ba.<e of middle ciiudal rays: the |>ores little or not at all branched. 

C>>lor in life (fiel<l No. 0:U(>4), light rosy olive, with violet shades, pale l)elow; center of each scale 
of back shining violet; dorsal nnldish flesh-color, its l>a.<e anteriorly yellowish olive; caudal flesh- 
color, rosy along the inlges; anal .similar, its e<lge light lavender gray: ventrals |)ale, shadeii with light 
orange; |>ectonil flt*sh-color, violaceous at base; snout violet, iris light yellow. A flesh-colore<l vio- 
la<*eous fish without color markings anywhere. Another sjK'cinien (field No. 03417 ) freshly dea<l, had 
the Ixxiy, head, and caudal light rosy; ventrals white: outer margin of spinous dorsal golden, the 
membranes with irregular golden areas: iM'<*toral and anal not distinctly colonel; iris yellow. 

Color in alcohol of tyin* (field >»<►. (KWIS), above dusky silvery, biis«*s of scales brown; sides and 
under parts silvery, with pale greenish-yellow tinge: top of head somewhat olivaceous, sides rusty 
silvery; axil of i>ectond <liL»<ky; fins all jiale or yellowish-white. 

This s})e<*ies rea<*he8 a length of alxait 2 fwt and is an important fcMnl-fish. 

Only 4 si)ei'imens were secure<I: 

! '^' : -^>'- 

I^K'ality. Final di*<|Mts-iti«>ii of HjKM'iuien. 


, Jmh* X. 

0301.H ' L»4 ! Honolulu Tyi>e. No. ri06«.0. l*. S. N. M. 

aT04<) I do CotvfM". .No. T17a. L. S. .Ir. I'niv. Mus. 

OiVMM : iio ('ot'v|K'. No. JTJl. l'. S. F. ('. 

(WIT I do <'olvi>f. No. V«M3. Ind. Tniv. Mus. 

27. Bowersia ulaula Jordan A Evermann, new s|M*cit*s. "r^n/Ai." 

Hea<l 3.6 in length; depth 3.8; eye 3.S in head; snout 3.8; maxillary 2.9; mandible 2.4; inter- 
orbital 3.6; preorbital 10; scides 8-<>8-14: D. x, 11; A. iii, 8; Br. 7; gillrakers 21 -^5. 

Bcnly long and slender, the dorsal outline in a low, gentle curve from tip of snout to l)asc»of craudal, 
the ventral outline but gently convex: head nunlerate, bluntly conic; snout rather short: mouth mod- 
erate, somewhat oblique, the jaws equal: I'-axillary moderate, slipping for Ita entire length under the 


narrow, thin prcN)rbitAl, its wi«Uh at tip 2.8 in eye; l^andfl of villifonn tet»th on vomer, ])alatine8, ttni^ie 
and jaws, those of ont<T wries in tlie hitter wan'ely enlHrj»e«i; oi)en*le en<lin}; in 2 ol)s«'un', flat spini»t», 
the sjuiee lx*twe<Mi them diHjply eniarginate but fille<i ]»y mem])rane; pn'ojK^n'h* rather «liHtinetly wr- 
ratts the teeth very short; eye rather hirj^, it-* lower Ininler in line with axip of ])o<ly; preorbital very 
narrow, mneh narn>werthan in H. rioffwrtv<; interorlntal sjiai'c* narrower than in the i)n»ee<linjr si»e<'ies, 
sliphtly eonvex; caudal iM^hnurh^ lonjr, its length from Iww of last <lorsal niy to first supiH>rting rays 
of caudal 1.7 in heml, its lea.**t wi<lth alxnit 2.1 in its least depth, which is 2.1 in its length; gillrakers 
rather numerous, clos«*-R't, the longt^st aljout 2.2 in eye; lins nnMlerately develoixMl, the dorsal contin- 
uous, without notch, its origin slightly lH»hind basec)f jKH'toral and etpially dist^mt l)etw(HMi tii>of sn<mt 
to l)ase (»f fifth or sixth dorsal ray; heiul 2 in distance from origin of anal to midille of last dorsal ray; 
first dorsal spine rather short, ulxMit 1.9 in length of second; fifth dorsal spine hmgest, its length t^jual 
to distaniv from tij* of snout to jnipil; last dorsal ray pro<luce<l, its length aKnit l.S tinu»s that of the 
preceding; anal similar to soft dorsal, its origin un<ler base of third dorsid ray; first anal sjiine very 
short, the thin! longi»st, its length 1.2 in diameter of eve; s<»ft anal similar to soft «lorsal, the last rav 
pro<hiccMl and of equal length with that of dorsal; t'audal <lenst»ly s<*aUNl and wi<lely forkisl, hilH»s tMjual, 
their length, measure<l from luise of first supfMirting rays (M|ualing that of head; ventrals nnt ]Niint4Ml, 
their tips not n^aching vent, their length l.fJ in hea<l: i»e<'toral l»»ntr, slightly falcate, its tip n^siching 
vent and much In^ytmd that of ventral, its length equaling that of hcail; siiilcs rather small, «'li»sely 
imbricate*!, deei»er than long, their e<lges finely ciliattnl; ctuvk and ojH'n'les scaleil, <i rows on c!uH»k; 
a large l)ony hum«*ral scale from whi<'!i extends as«»rics«»f nuMlitied s<*ah's t<i najM', and in fn»nt of 
which is a jMitch of onlinary s<*ales; lateral line complete and well ilevelojHfl, lM>ginning jit lower t^lge 
of humeral scale and following contour of Imck to base of middle cau<lal rays, thetulx's little ])ranche<l. 

(\)lor in alcohol, browni.»*h or purplish oliva<*<H)us alnive, {Kiler on side; under parts ni»arly ))lain 
white; each scale of Iwck and ui>|K'r juirt of side with a darker bniwn s|K)t, thes*» forming indistinct rows, 
alwMit 6al)ove lateral line; side U'low lateral line with Icssdi.^tinct Iiorizont4il lim?s; up]MT parts <»f lu'ad 
oliva<'<'ous brown, lower parts jwiler, spines of dorsiil fin purplish, the membranes white, i)ur))lish at tii>s; 
soft <lorsid with rays whitish, membranes purplish; caudal slightly dusky, other fins plain whitish. 

This s|K»cies is relate<l to />. riohwciij*, inmi wliicli it differs chiefly in tlu' shorter snout, larger 
ey«", shorter maxillary, shorter mandible, narrower intcn)rbital siiac<>, de<'i<le«lly smaller scales, more 
numerous gillrakers, and more jK»stiTior ins(>rtion of dorsal fin. Only one siN^'imen known, ty{»e 
No. ry(Hm, U.^S. N. M. (fiel«l No. 04104). I4.2."> indies long, from llilo, Hawaii Island. 

28. Etelis evurus Jordan ^ KvtTmann, new siMH'i<*s. 

Ilea*! 3.2 in length; <lepth :\M; eye.'i in hea<l; snout :5.1»: niaxillary 2.2; interorbital :>.f>; D. x, 11; 
A. Ill, 8; s<'ales r>-r>()-ll; \W. li; gillrakers 15 • IJ, l<in^<*st alntut 2 in eye. 

I^sly rather long, taiK'ring, moilerately e<»nipress<'4l; <Iorsal outline slightly convi'X, ventral «»ut- 
line nearly straight; hea<l «'onsi<lerably longer than deep, eompress<Nl, huImmuhc; snout bluntly i»ointe<1. 
h»ss than eye, (Mjual to jM)rtion of eye anterior t«) jHisterior iil^r nf pupil; niouth large, oblique; snuill 
Imnds of villifonn ti-eth «»n vomer, |»alatin<'S, and anterior part of eiu'h jaw; a single niw of small, 
wide-set, slendi*r canine X*>v\]\ on th«* outer e^lifi* of inich jaw, tfms*- in np|M'r jaw slightlv largiT and 
rnort* wi<le-s<'t; a single larger canine t«H»th on the sitle of ea«'h jaw in front, t!ios«> in the up|KT jaw the 
larger; maxillary extending to middle of pupil; eyi* very la rg«', its lower edge slightly U'low axis of 
Isnly; jinM»|M»rele finely wrrate; ojx'rde with 2 broiul, flat s|»ines, not pnHluced, the up|K*r rather 
ol>s<'ure; fins UMMlerately d«'velo|M»<l; origin of spinous florsal hliyhtly {Misterior to li;i.>^Mif jHH-toral, its 
distance from tip of snout (Mpialin^ tliat Ut Inu^' of sixth dorsal my; dorwil fin deeply not(>he«I, alm(»st 
divid*»d; first dorsal spine short, its length but hliirhtly ^n"at«-r than fliameter of \m\n\; thin! dorsal 
spine longivt, 2.1 in hea«l; ninth spine short, its liMiirth 2.75 in tliinl; soft dorsjU not elevated, the ravs 
alNHit (H]ual, the last 1.75 in third spine; anal similar to H»ft ilorsal. the first spine very short, the third 
al)out l.S in third dorsal spine, Isist anal ray aUmt eipial in la-t ilorsal ray; caudal deeply notche«I, the 
IoIh's much priMln<-e<l, the up|»i-r the |f»mrer. its rayr- jrreatly t'xceiilinL' lenu'th of !ira<l, or alMuit 2.4 in 
IwMly; ventrals long, but not n*Jicliing \i'nt by a distan<-»' i-qualin'^' half diameter of pnpil. tln'ir lengtli 
1.5 in beat!; iH'ctoral long, n-arhing vent, the npjii-r rays somfwliat pr<Hhie«H|, their leiijrth 1.2 in hea*l- 
S4'ales UKNlerate, firm, i-overinj: InMly, najH', ojHTrN-s, anil bn-a^t; a largi* hnniend s<'ale; lateral line 
l)eginning at lower e<l^r«• of huiii**ral s<-3ile and fo!l«»win;r contour of Isick to I>:is4' of caudal fin. 

(%»lor ill life, of a siM-^-imen (field .N'«». 0."rlHl ) 1 1 itichi-y lonir, brilliant n»>e-nH|, the side from level 
of eye abruptly silver, with rosysha4les; snout, jaw-. e\«-. jiinl inslfle of iii«>uth nnl; finn all ruse-iXilor, 



the dorsal and caiulal bri^rht; ventrals and anal pale, the former washed with red on center; axil 
pale pink; pe<-toral pale ro<«y. 

Color in alcohol, uniform yellowish white, paler below; fins all pale yellowish white, the caudal 
UAyes somewhat dark. 

This species is relate<l to Ett'li^i ot^nhttun of the West Indies, from which it differs in the somewhat 
larger scales, much lonj^r caudal lolx^s (9.5 times leuj^h of middle rays instead of 4 times, as in E. 
onihituA)^ and larjjer eye. From E. mrhnnrnbiH C'uvier A Valenciennes, fn>m Isle de France, it seems 
to differ in having only Irt instead of 'JXS scales in a transverse series and in the coloration. 

This gpe<Mes, one of the handsi»mest of all Hawaiian fishes, is thus far known only from Hilo, 
Hawaii, in the market of which we ol>taim»<l 13 fine examples, measurements of which are given in 
the f<»llowing table: 

Length . 


Finjil disiMwition of >i»*^'i- 




Final disposition of .«iiet'i- 








Tvpt'. No. oOiU?2. r. S. .N. M. 




Cotype. No. 24*224. Ac. Nat. 


13. -25 

<ViIvih:. .No. •27'22. T. S. F. C. 

S<n.. Phila. 




(^ot\lK'.No.;iOlV«.r.S. N. M. 


12. 75 

Colvf»e. No. 3a5S. Field Col. 




CotyjH'. No. 7174. L. S. Jr. 


Tniv. Mu.t. 



.. .do 

Cotyi»e. No. »<14. Ind. I'niv. 






15. 5 


Col v|K\No.l701.Bi>hoi> Mils. 



('otvi»e.No.l498.Cal.Ao.S< i. 


12. 7.> 


<'otv|K-. No. . M. r. Z. 


12 6 



Cotvpe. No. *229t'.. Am. Mas. 



Nat. ni>t. 

Family kYPHOSID.€. The Rudder- fishes. 

29. Sectator azureus Jonlan Jk Evermann, new species. 

Hea<l 4 in lenjiTth; <lepth .S; eye 5 in hi*ad; snout :i.»>5; maxillary 4; interorbital 2.4; D. xi, 15; 
A. Ill, i:^; s<"ak»s 14-81-20. 

B<Hly clongjitc, ovoid, greatest depth alMDut at tip of |>e<*toral; lK»a<l slij^htly longer than deep, 
compresse*!; snout very bluntly ronvex; jawsalx>ut c<|ual. maxillary not reaching front of eye; mouth 
small, horizontal; teeth very small, comj>ressi*<l, in a single st»ries in t^ch jaw; minute villifonn teeth 
cm vomer, palatines and tongue; tongue bn)ad, rounde<l and fnv in front: pn'ojH^n-le entire, posterior 
e<lge very oblique: lower eilge of eye on a line with up|>er ba.»«e of |)ectonil, |M>jiterior margin well in 
front of middle of h<»a<l: interorbital bn»atl, strongly convex, a deep gnxive in front of eye to nostril; 
caudal ihmIiuicIc rather long, l.^> in hea«l: oriirin of spinous <lorsal slightly in front of ba^H? of ventrals, 
well l»ehiud i>e<'toral, its <listan<-c fn>m tij> of snout slightly grt»ater than <lepth of Ixxly; longest dorsal 
spine 3 in h*'ad. liL^^t dorsal ray eh^ngat*', U'lm: one-fourth longiT than other rays, its length 3.4 in heail; 
thinl anal spine longest, 4.9 in htiid: first anal ray longt»st,3.4 in head; ba.*H.^ of anal l.S in baseof dorsal; 
cau<lal <leeply forke<l, lower IoIh» the longer, 3.5 iu ImmIv; jKN'toral short, slightly longer than ventrals, 
1.8 in heacl, the spine more than half length of l<»ng*»st my: s4-:iles cycloi<l, present on hea<l except on 
jaws and in fn»nt <»f eye, very minute on all the fins except ventrals; lateral line concurrent with dorsal 
outline; peritoneum dark gray. 

Color in life, *lark steel-blue, lH*<*oming }»aler UOow; a <lefinite deep blue stripe from snout l)elow 
eye widening on ojien'le, and thentv .»*traight to ci^nter of l>a.«*e of caudal; Ijelow it a narn)w bright 
golden strii>e fn>m angle <>f mouth to lower jwrt of caudal, and then a fainter blue stripe Ik?1ow this; 
a blue stri|>e from eye to upjHT jiart of gill-o|>ening, inters|>a(V sha<le«l with gret*n; a dc»ep blue stripe 
fr«)m ujUK'r j»art of eye along each si<le of l>ack to l» of upj>t»r cau<lal lolxs upi>er fins <lusky golden 
or olivace<^us; ventrals yellow: anal and lower IoIh» of caudal dirty g«)l<len: i»ectoral translucent. 

Color in alcohol, deep st«vl gray, brown alntve, ea<'h S4'ale with a very j»ale sp<jt, the etige pale; 
lower surfa<v whitish silvery; a pale streak of gray l>ehind eye to t^lge of opt^ri'le: dorsal fin gray 
bniwn like the l>ack: <-siudal an<l i>eetoral whiti.<h: insi<le of ventrals ilu.'jky orange; ventrals and anal 
dusky: inside of i>ectoral blaekish brown. 

Ty|>e, No. 5<)<UM, U. S. N. M. (field No. 0:>:i6:5), a sjiei-imen 15.25 inches long, taken off the shore 
near Heeia, Oahu Island. 

This species must Ik* very rare, l)eing unknown to the fishermen and only tlie single specimen 
having been obtained by us. 



Family MULLID^. The Surmullets. 
30. ICulloides flammeuB Jordan & Evennann, new species. 

Head 3.6 in leng:th; depth 4; eye 4.3 in head; snout 2.25; interorbital 3.5; maxillary 2.6; 
mandible 2.1; shortest distance from eye to upper edjure of maxillary 1 in eye; D. vii-fl, longest dorsal 
spine 1.75 in head, longest dorsal ray 2.6; A. 7, longest ray 2.7; scales 3-41-6; pectoral 1.5; ventral 1.4. 

Body oblong, not much compressed; head heavy, broad, the interorbital space broa<l and slightly 
convex; snout rather long and pointed, not abruj)tly decurve<l; mouth rather large, somewhat 
oblique, the lower jaw but slightly included; maxillary bmaci, slipping for most of its length under 
the thin preorbital, its tip not reaching orbit by diameter of pupil; eye rather large, high, slightly 
posterior; gillrakers 18 -f 7, the longest alxiut 2 in eye, serrate; oi)ercular spine obscure in adult, more 
plainly developed in the young; origin of dorsal a little nearer i)Osterior bast* of soft dorsal than tip of 
snout; distance between dorsals considerably less than snout, al)out 2.6 in hea<l; anal similar to soft 
dorsal, its origin somewhat more posterior; ventrals rather long, reat^hing slightly l)eyond tip of 
liectoral; caudal deeply forked, the lobes equal, about 1.2 in hea<l. 

Color in life (field No. 0,'U59), bright rose-red, with 5 broad crossbands of darker clear rose, 
which vanihhes very soon after death; a very faint yellow lateral streak, with yellow sha<les on scales 
below; lower side of head rose, snout and lips very red; 2 wavy golden streaks frouj Ih»1ow eye to 
angle of mouth, lower conspicuous; first dorsal clear red; second dorsal dei»p red on the lower half, 
fading above; caudal deep red at ba«e, fading outward; anal pink, pec^toral light yellow; ventral 
creamy red; barl>els re<l, paler toward tip; iris silvery. 

A color note on si)ecimen8, field Nos. 03054 and 03055, says that they were rosy in life. 

Color in alcohol, pale dirty olivaceous above, yellowish white on sides and belly; hea<l- yellowish 
olive above, pale on cheek and below; a yellowish band from snout under eye; fins all colorless, the 
spinous dorsal slightly dusky, all with slight yellowish tinge; ventrals with the middle membranes 
blackish. Smaller examples show c(msiderable rosy on the sides, indicating that the fish in life was 
probably red or rosy in color. 

This species somewhat resembles Mnlloides auriflnmma, from which it differs in the smaller eye, 
larger, more oblique mouth, longer maxillary, the longer, k»ss <lecurve<l, more i>ointed snout, and fewer 
gillrakers. It lx?ars some resemblance to M. pfingeri, but has the eye larger and the snout longer and 
more pointed. Compared with M. wimorusis, it has a much largt»r and more oblique mouth, and a 
considerably longer maxillary, as well as a <lifferent coloration. It does not agree with any of the 
plates of Day, Oiinther, or lileeker, nor with any current des<.^rii)tions. In life its Imndtnl coloration 
gives it a very handsome appearance. It is found in deeper water than most of the other species. 

M. flammeun seems to lx» fairly abundant and is represented in our collet'tions by the following 9 
specimens, the first of which is taken as the ty|)e and all the others as cotyiws: 









Final disposition <»f Hpcoi- 

Kiiilua ... 






Type, No. 50665. U. 8. N. M . 
(k)type. No.1702. Bishop Mns. 
Cotype. No. 7475. L. S. Jr. 

Tniv. Mils. 

Cotype. No. . M. C. Z. 

(^)tvpt^ No.27'A U. S. F.C. 
Cotypi'. No. 2297. Am. Mus. 

Nat. Hist. 






! 113931 






T^w.uiitv ■ Final dispo«<ition of HiKH'i- 
*^^""-- men. 

Hilo < 'otyi»c. No. 24225, Ac. 

I Nat. .S<i. I'liila. 

do — I Coty^K*. No. 9S15. Ind. 

I'niv. Mus. 
d«» — CotyiH-. N«». 39:»;^. Field 
Col. MuH. 

31. Pseudupeneus chiyBonexnus Jordan & Kvermann, new S]>e(!ies. 

Head 2.8 in length; depth 3.4; eye 5.3 in he^id; snout 1.7; interorbital 3.5; maxillary 2.3; D. 
V£ii-9; A. I, 7; scales 3-1^0-7. 

Body slender, not greatly compressed, the back gently and rather unifonrdy elevated from tij) of 
snout to dorsal ; ventral outline slightly convex; heatl mwlerate; snout long; bluntly pointed; mouth 
moderate, slightly oblique, the lower jaw inclu<led; maxillary broad at tip, falling short of vertical of 
orbit by diameter of pupil; interorbital space convex; eye small, in posterior half of hea<l; teeth rather 
large, in a single band in each jaw; barbels long, 1.2 in head, rt»aching nearly to l) of ventrals; 
opercular spiue small; fins rather large; third dorsal spine longest, 1.5 in head, or equal to distance 



from tip of snout to muldle of pupil, thin! ray longest, 3.2 in head; ba^e of spinous dorsal 1.4 in third 
spine; ba.^ of soft dorsal 1.4 in longest spine; origin of spinous dorsal nearer laet dorsal ray than tip of 
snout by longitudinal diameter of pupil; distance between dorsals 1.5 in eye; length of caudal peduncle 
1.5 in head; i>ectoral long, pointeil, slightly falcate^ 1.4; ventrals slightly longer, 1.3; last anal ray 2.9, 
equal to base of fin; caudal shallowly forked, lobes 1.3 in hea<l, nuddle rays 2.75 in upper lobe; scales 
finely ctenoid and i)l)s<'urely dendritic; lateral line concurrent with the l)ack, the pores with few 
branches, the number usually not exceeding 5 or 6; 2 scales Wtween the dorsals, 8 on dorsal side of 
caudal j)eduncle; i)eritoneum srjmewhat silvery. 

Color when fresh, deep scarlet re<l, es}XH*ially a shade from snout through eye toward tail; first 
dorsal plain scarlet, second paler golden with oblique stripes of s<'arlet and yellow cMlge; caudal orange; 
reddish at base, yellowish at tip; anal like si'cond dorsal; jHM-toral i)ale orange; ventrals deep red; 
bar>>els bright yellow; iris re<l. In life, a i)ale streak backward from eye to miildle of side parallel 
with back; side with 2 blotches of deep red; a row of dark spots along bases of lx)th <lorsals; young of 
3 inches, from the rock pools, in life, dark olive green above with a dark olive streak along lateral line 
and 3 dark sha<les under first dorsal, second dorsal, and l>ack of caudal i)eduncle; tip of first dorsal 
cherry-red, edged with white; second dorsal and caudal translucent, scarcely reddish; ventrals and 
anal bright cherr\^-re<l, former mesially dusky; l)arl)els golden. 

Color in alcohol, pale yellowish; each wale below dorsal with brownish eiigings, generally most 
distinct in young and often entirely disappearing with age; a serii^s of smaller obscure spots along 
median line from oi)ercle to tip of pectoral; sides anil under parts with faint traces of rosy. 

This species may be known by the series of dusky blotches along each side of the dorsal fin and 
by the simi»le structure of the lateral line. In life it is at once known by its golden barbels. 

The al)ove description base<l upon a specimen (field No. 03929), 8 inching long, obtaine<l at Honolulu, 
in 1898, by Dr. Woo<l. We have examined 4 other snecimens of approximately the same size obtained 
at the same time, and numerous examples coUet^ted by us at Honolulu and Hilo. 

The following is our list of specimens of this si)e(*ies: 







o:w«6 . 

039S7 ' 

089SS I 

.'>. '25 
(J. T.'i 




Final di.Mjiosition of spe<i- 

Inrht X. 



TviK-. .S'o.50666.r.S. N. M. 

1. 5 



' Hil(. 




»). 5 





s. 5 







5. 25 




do . 

di» . 

do . 

do . 

do . 

do . 

Cotypo, No. 7476. L. S. Jr. 

I'niv. Mus. 
Cotyiif. No. 7170, L. S. .Ir. 

I'niv. Mu.K. 
CotviK'. No. 2724. r. S. F. ('. 
t'otvfif. No.'27'25. r. S. F. <:. 
CotviK?. No. 170;J. Bi.^hop ' 

Cotvpo. No. . M. C. Z. 












Length. , Ltntality. 






OH :, 
'«.5.5 ;. 

« 5. .5 1 . 
<« 5. 5 



do . 

do . 

do . 



B « « * * "-« * ' » « ■ * 





Final dis|M».sition of spteei- 

Cotype. No. 229^. Am. 

Mus. Nat. Hist. 
C'otvfK-. No. 21*226. Ae. 
Nut. S<'i. I*hila. 
, Cotype. No. 9S16. Ind. 

Tniv. Mus. 
I Cotvi.e. No. 3955. Field 
I Col. Mu.'*. 
Cotvpe, No. 1499.Cal. A<-. 

Cotvr»e, No. [i0676. V. S. 
N". M. 

« Collected by Dr. W<kk1. 1h9h. 

'>f>3llecte<l by Jordan & Snvder in 1900. 

32. XJpeneus arge Jonlan & Kvermann, new si>eties. ''Weke^^ or ^^Weke Puco^ 

Head 3.75 in length; <lepth 4.1; eye 5 in head; snout 2.25; interorbital 3; maxillary 2.3; shortest 
distance Ix^tween maxillary and eye 1.25 in longitudinal diameter of eye; D. vni-9, second spine 1.5 
in head; A. ii, <>, longest anal ray 1.9 in head; j>ectoral 1.5; ventrals 1.45; scales ;i-40-7. 

Brnly oblonir, compressinl, deei)est through the anterior of the spinous <lorsal; head moder- 
ate, compresse<l, profile an'he<l from origin of the spinous <lorsal to tij) of snout, steepest on snout; 
snout bluntly roun<le<l; lower jaw inclu<le<l; mouth mcMlerate, slightly oblique; tongue short, rounde<l 
anteriorly, not broa^l nor thick, and not free; teeth in villiform bands on each jaw and on vomer 
and jmlatines; maxillary mfnlerate, reaching anterior e<lge of eye, mo<lerately broad and sheathed 
for more than half of its length; eye rather small, high, mt»dian, adipose eyelid well developeil; 
barbels not reaching e<lge of gill-oj>ening; i»seudobranchia» well developed; gillrakers 16 r6, finely 



eerrate, last 5 or 6 on longer limb very blunt and short, pupil of eye containel 1.5 in longest; spinous 
dorsal 1.5 in depth, first 2 spines even, longer than the others and longer than l)aHe; distance from 
snout to origin of spinous dorsal one-thinl distance from snout to last scale on caudal; distance l)etween 
dorsals slightly less than base of soft dorsal; soft dorsal slightly concave; caudal deeply forke<l, upper 
lol)e longer; anal similar to soft dorsal, in8erte<l slightly liehiiid the latter; ventrals reaching slightly 
Ixn'ond pectoral, rays of pe<^toral slightly the longt'r; lateral line concurrent with dorsal outline; scales 
large, finely ctenoid; entire Ixxly and head scaly. 

Color in life, palegretm, changing to white below; e<lges of stales on Iwck and down to lateral 
line purplish brown, giving the appearance of 3 rather distinct strii)e8 of purplish brown, with 
greenish centers on the scales; side with 2 broad yellow stripes, the up|)er l>eginning on opercle at 
level of eye and running to caudal just above lat<»ral line, the latter l)eing croese<l under soft dorsal; 
second banning on base of pectoral and running to base of <'audal just Inflow lateral line, this striix; 
less distinct and narrowing ])OSteriorly; oi)en!le bright rosy; top of head dusky; cheek whiti* with 
some rosy; lower jaw white; barbels yellow; dorsal fins pale, each crosse<l by 2 or 3 brownish nwy 
Imrs; caudal white, upper lol)e with 4 broad brownish red bars numing downward and l)ackward, 1 
at base narrow; lower lobt» with similar but much broader l)Iack bars nmning upward and l)ackward, 
2 of them more distinct than the others; 2 longish <lark spots on inner rays; anal, ventrals, and 
pectoral pale, ventrals rather pale yellowish; iris yellowish, pink al)ove. 

Color in aWhol, al)ove, bluish olivaceous, the side becoming lighter, almost white on belly; 
boniers of scales dusky; first dorsal spine with 3 or 4 dark sjwts, and the upi)er ix>sterior edge (»f 
membranes with dark si)ots; soft dorsal with 3 dark spots on anterior eilge and similar spots on 
upper part of fin; caudal fin with dark Ivands, upi)er lobe with alnnit fi, those on lower lobe 4, mu(rh 
broader; other fins jmle. 

This species restMubles rperuna rittatun (Forskal), dewrilKnl from Djidda, Arabia, but the latter 
has the l>elly abruptly deep yellow in life. 

This is an abundant and imj^ortant fo<Kl fish at H(molulu, where we obtaininl 10 s{)ecimens and 
where 4 others wen^ collecttnl by Dr. Jenkins in 1889. It is e<|ually common at Hilo and in IVarl 
IIarlK)r. It lives in shallow water along (piiet shores, and is known as ** Weke" or ** Weke I*uco.'* 

The following is our list <»f specimens: 






lA'iiKtli. , L(N*ality. 

Final dis^KMiiioii of 




9. 2f> 

12. .•> 

Honolulu. Type, No. 50007. U.S.N.M. 

<io ' ('olypo. No. 7477, L. S. Jr. 

ITniv. Mu>«. 
CotyiH'. No. 2726, U.S. F.C. 
< -otyiH'. No. 1701. BiHliop 

(V)tyiK', No. M. C. Z. 

(^otyiH'. No.2299,Ani.MuH. 

Ntit. HiHt. 
( \»tyi>e, No. 24227. Ac. Nat. 
Sri. Fhila. 

do . 

do . 

.do . 
.do . 






Final disiMtsition of 


10. -25 


rotyiH'. No.SOTvl. FieldCol. 

M UN. 








<:otyiM'. No. XiHX), <'al. Ar. 








10. :» 



npnieoidrt* viWUm, StreetM, Bull. U. S. Nat. MiiH., No. 7, 71. 1H77 (Honolulu): not of F.)rskjll. 

Family F()MACENTR1I)/K. The Demoiselles. 

33. GlyphiBOdon sindonis Jordan Jk KA'ermann, new s|KH'i(»s. 

Hea<l 8.5 in len^^th; depth 1.75; eye :{.4 in heiul; snout 8.5; maxillary 8.4; interorbital 2.8; I). 
XII, 19; A. II, 15; s<-ales 4-28-9, 22 pores. 

Body short and det^p, dorsal outline evenly arched from tip of snout to soft dorsjil; lu»ad (UH»i)er 
than long, comi)resstNl; snout short and ronic; mouth small, horizontal, lower jaw slij^htly shorter; 
maxillary reachinj^ to anterior tHlge of orbit; a sinj^le row of small, nithor blunt, slij^htly compresst^l 
teeth on esx'h jaw; preopercle entin», opercle ending in 2 small flat spines, njiper very small ami 
obscun*; eye anterior, hijfli, its lower e<lj?t» above upixT base of |HH'tonil; interorbital broad, steep and 
convex; fins larjre, orijjin of dorsal oviT l)a.*«M)f ventrals, its distiu ice from tip of snout (Mpial to di.«- 
taiK'e from l)asi> of last niy to tip of upin^r mmlal loUs spines strong? and lonjr, first 0.7 of fourth, which 
is 1.9 in head and of same length as following spines; mid*lle dorsal rays produceil, longest ray 1.25 in 


head; anal pimilar to soft dorsal, longest ray 1.25 in head, Seconal spine longiest, 2 in hea^l; caudal 
forkcnl, upi>er Io]>e the lonj^er; wntrals n*a<hinp pai*t vent, outer rays longest, al>out espial to head; 
pectoral broa<l, upjKT rays longest, ecpial to hea<l; scales large, ctenoid, covering entire lx)dy and 
head excvi)t lower jaw and snout anterior to eye; lower limb of preopercle scaknl; large 8<»le8 cover- 
ing nearly all of dorsal spines, smaller scales cohering as much of soft dorsal and anal and nearly all 
of caudal; very minute scales on ba<e of |>ectoral, none on rays of ventrals; lateral line concurrent with 
<lorsal outline, on 22 scales, ending '^ rows of scales short of posterior l>ase of dorsal, then dropping 
3 rows of scales and continuing ol)scurely on mi<ldle of caudal jKiduncle to l)a«e of caudal fin. 

Color in alcohol, uniform very dark brown, nearly black; 2 narrow wavy Iwinds of white on side, 
first beginning alx)Ut under fourth dorsal spine and extending under alnrnt middle of i)ectoral, thence 
curving slightly backwanl toward vent, rather indistinct l)elow pectoral; second band Ixyinning under 
last dorsal spine and first ray, extending towanl middle of anal, rather ol)scure, indistinct for 2 or 3 
scales before reaching anal; tins all black, pectoral slightly lighter than others; a large black ocellated 
spot with a narrow white border on back and lower part of soft dorsal, larger than eye, just back of 
last white bar. 

This sjKJt'ies agrees with tyi>i(«l Glyphisodon in all respects except that none of the teeth appears 
to be emarginate. It agrees with Chrtfitlptera in the entire preopercle and preorbital and naked snout, 
but differs from the type of that genus in having the teeth in a single serit»s. 

The above dest'ription base<i on the tyi)e. No. 50669, U. S. N. M. (field No. 04524), a specimen 3.75 
inches long, from Honolulu. One other specimen obtaine<l. It is taken as a cotype and is No. 2727, 
U. 8. F. C. reserve series (field No. 03732), a specimen 2.75 inches long, from Kailua, where it was 
first discovered by Michitaro 8indo, for whom the species is nameil. 

34. Pomacentrus jenkinsi .Ionian <fe Kvermann, new species. 

Head 3.4 in length; depth 1.8; eye 3.3 in hea<l; snout 4; maxillary 8.2; interorbital 2.75; D. xiii, 
16; A. II, 13; s<ales 4-29-11; Br. 4. 

Body ovate, deep, compresse<l, dorsal outline rather steep, evenly cur\'eti from tip of snout to soft 
dorsal, following e<lge of scales on spinoiLs dorsal; head deeix*r than long, compressed subconic; snout 
bluntly conic, jawseijual; maxillary reaching anterior edge of eye; mouth small, horizontal; a single 
row of close-set, incis<^)r teeth in each jaw; jx>Hterior e<lgeof pnM)i)ercle roughly serrate; opercle ending 
in 2 short flat spines, the up[)er very ol)scu.*e; interorlntal wide, strongly convex; fins rather lai^ge; 
origin of dorsal over ventral, origin of each e<|ually distant from tip of snout; first 2 or 3 dorsal spines 
shorter than others; others alH)ut of e^juai length, shorter than the longest dorsal ray.s, the median rays 
being longi»st, 1.5 in hea<l; caudal forke<l, lol)es rounded, up|K^r the longer; anal rounded, longest ray 
1.5 in head, second spine rather stout an«l strong, 2.2 in hea<i; ventrals long, reaching vent, 1.1 in head; 
pectoral broad, upjier rays the longer, 1.2 in head; scales large, finely ctenoid; Ixxlyand head, except 
lower jaw and snout, scale<l, scales on to[> of hea^l small; bases of all the fins except ventrals well 
covere<l with fine scales, those on si)inous dorsal larger; lateral line concurrent with dorsal outline to a 
line under base of third or fourth dorsal ray, where it drops 3 rows of scales to middle of caudal 
peduncle, whence it continues to base of caudal fin, the detache<l jKjrtion little developed. 

Color in life, ground color dark drab; central jntrtion of s(*aies oliva<'eous, each one with black on 
lower part of posterior e<lge fonning vertical bands on Ixxly; axil black; outer Iwrder of dorsal fin, 
above scale<l i>art, black; pectoral dusky olivaceous, black at base; ventral and anal black; caudal 
dusky with posterior l)order lighter; iris bright yellow. 

Color in alcohol, dark brown, edges of scales darker; a dark strii>e on upj^er eflge of membranes 
of spinous dorsal, broa<lestand most distinct anteriorly; rest of dorsal, and caudal and pectoral dark 
brownish; ventrals and anal dark, almost black; a black blotch at upper base of iwctoral, continuous 
with the black axil. 

This is a very abundant si)ecies among the Hawaiian Islands. Numerous specimens were obtained! 
at Honolulu in 1889 by Dr. Jenkins, and others by Dr. Woo<l in 1898 and Dr. Jordan in 1900. Our 
own collections, made in 1901, contain numerous specimens, the localities represente<l l>eing Honolulu, 
Hilo, and Kailua. 

The alx)ve description is based chielly upon a si>ecimen (field No. 04526) 4.8 inches long, obtained 
by us at Honolulu. 



The field niiinl)ers and lengths of a few of our specimenp are jnven in the following? table: 

^.*„>0 j U.„K.h. 


Filial (1iHp<)Mitloii of spi'oiineii. 
























1 1 




do — 


do — 



Kailua ... 

do — 


No. 7479, L. S. Jr. I'niv. Muh. 
No. 74SO. L. S. .Ir. Univ. Mas. 
No. ;J«V9. Kield <'•»!. Mus. 
No. 50fi71. r. S. N. M. 
No. 2728, r. S. F. (\ 
No. 2?29. V. S. F. C. 
No. 1705. Bishop MUR. 

No. . M. C. Z. 

No. 24228, Ac. Nat. Sci. IMiila. 

No. 5Wi70. U. S. N. M. 

No. 2300, .\m. Miu». Nat. Hi«l. 

Ponuirfntrus niffricantt, Quoy «i: Gaiinard, Voyage Uranie, Z(K)1., 399, 1821 (Sandwich l}iilandt<): CuviorA: ValoniMcnufs. Nut. PoiHH., V, 125, 1830; (fiinthcr, Cat , iv, 34, 1^2 (Sandwich Islands) ; not JlnlfH-entnn* nifjricaji^ 1^ic<^|hMc, 
HiHt. Nat. Poi.sM.. iv, 332 and 3<i7, 18(Ki, Icwality unknown, collwttHi by Coramcrson. 

Eupomnrenlruis manjinntw .laixkiiw, Bull. U.S. Fish C<»min. for 1899 (.Tunc 8, 1901), 391, 11^.5, Honolulu (Type, No.4*.)700, 
U. 8. N. M., Coll. O. 1*. Jenkins); not PoniacentruH mnrginatws Riippell. 

Family LABRII)^. The Wrasse- fishes. 
35. Lepidoplois strophodes Jordan &. Kverniann, new Pi)eeies. 

Head 2.75 in length; depth 2.75; eye 4.65 in head; snout 3.25; mouth 3.1; interorbital 4; I), xii, 
10; A. Ill, 12; scales 7-34-13. 

Bo<ly oblonj?, compressed; head longer than deep; upper and lower protiU*H evenly and slightly 
convex; snout long, pointe<l, rounded alwve; jaws produce<l, i)ointetl, alx^ut etpial; mouth large, 
maxillary reaching beyond front of eye; teeth strong, forming a sharp cutting e^lge on sides of jaws, 
front of each jaw with 4 large mnines; eye rather large, anterior, high in head; i>osterior margin of 
preopercle very finely emarginate; interorbital S|)ace rather l)road, convex; nostrils small, anterior in 
short tube; dorsal spines [mngent, longest 3 in head, last 3.5; thini anal spine longest. 2.8; thin! anal ray 
1.9in head; i^ectoral rounde<l, 1.7; ventrals pointtnl, 1.4; caudal broatl at Iwse, tnincati*; i^udal iHMluncle 
broad, compressetl, its depth 2; scales large, thin, those on front of dorsal, along its l)ase and that of 
anal, small; lateral line concurrent with back, sloping down at (audal, then running straight to iti^ base. 

Color in life, pale rosy white; upper j)art*< of the snout, nape, ami side to l>a.»<e of alnjut ninth 
dorsal spine, lemon-yellow, extending down on side to level of ui)per tnlge of imjul; side of hem! 
very i)ale rosy, 2 irregular broken lines of wine-(!olore<l spot« aitross snout and through eye to 
posterior edge of ojwrcle, a similar row of 4 oblong spot« from angle of mouth downwanl and 
backward to e<lge of oiKjrcle; cheek and side of lower jaw with numen)us small irregularly i)laced 
orange spotj^; side with about 16 brighter rosy longituduial linej:^, those above less <listinct on 
account of the deeper rosy ground color, those Ijelow more distinct, the ground color iMMng more 
white; side lu^tween anal and sf)ft <lorsal (ins with a broad sooty ])lack sj»ot extending irregularly 
upon l)oth fins and fading out upon IxKly anteriorly, the iM)sterior e<lge l)eing nearly vertical and 
well defined; caudal iKxluncle and base of the caudal fin whitish, with a slight tinge of rosy, a pale 
rosy ban<l se]>arating this from the black lateral area; region in front an<l below the j>e(!toral with 
about 4 series of small reddish brown s[>ots; jHiCtoral region and the under i)artH somewhat bluish; 
dorsal tin rich lemon-yellow, the tii>s of the soft rays whitish, and a small, n>und, l>lack si)ot on 
middle of membrane of second spine; l>a«e of soft rays and last ilorsal 8[)ines rosy from intrusion 
of the rosy wash on side of body; last dorsal rays sooty black at the l»ase from extension of the 
black si)ot on the side; caudal i)ale lemon-yellow; anal pale rosy in wnter, lemon on spines and 
along tip of fin, l>ase of fin sooty black from intnision of black spot on side of the Ixnly, the black 
extending farthest down on the interradial membranes; iKH^toral very pale rosy; ventrals i>ale 
rosy, the membranes bluish, the tip of second ray blackish. 

Color in alcohol (field No. 04291), gray-brown, gradually <larker ixwteriorly; space l)etween soft 
dorsal and anal abruptly black, the color extending forward in darker streaks along the rows of s(rales 
and forming a large black blotch on soft dorsal and anal; top of head and space l>efore dorsal abruptly 


pale; posterior part of caudal pe^lunole also a>)niptly jiale; a black blotch on dorsal Ijetween second 
and thin! ppines, not involvinir thinl an«l fourth, sa* in L. hlluuuhitujt: dorsal and caudal rjtherwi«e 
pale; a jwile blotch at l>a.«*e of posterior dorsal rays; side with nariDw dark brown longitmlinal lines, 
coalescing; p<»steriorly with the black blotch; 2 narrow bn^wn streaks from lip to front of eye, then 
liack across side of head al)Ove; eiljre<l with narrow, <larker, wavy Hues; a wavy streak from comer 
of mouth towanl base of pectijral, lower side of head with small brown spot.s or blotches; ventral fin 
mostlv duskv. 

This species is very close to Ij*indophis hUunuhitHs^ differing chiefly in the dark zone on posterior 
part of body and in the smaller size <if the dorsal spot. Our si»e<*imens an* all young, but we have the 
young of L. hihnuilntujt si-arcely larger and showing the markings of the adult. 

Our collection contains the following specimens of this species, all from Honolulu, where it is not 

Ti 1 1 

\y: Ix'tiifth. Locality. Final 4lL»iMi^itioii i»f >in.-<.'imeuj 

lUfThf J*. 

041*91 4. 7 Hoiifilulu T\\H\ No. .T0r;72. U. S. .S*. M. 

iKrtJO 3.H «lo r4.tviH\ No. jt:)!*. r. <. F. ('. 

ttt=«2 4..S <lo ('.»fyjK?. No. 74^^I. L. S. .Ir. I'liiv. Mus. 

WJtRi S. 75 do i'ot'vpt'. No. 17i»i. Bi-ho|» Miis. 

0121W 3.7.S do <V»l'v[K-. No. 3*wi. Field To). Mii>. 

yEBBICULUS Jordan ft Eyennann, new genus. 

I >moi/iM Jon Ian 6c Evemiann, new gi*nus of J/ihritl:r {Mimpiiutun). 

Body elongate, subfusiform, compn*sse<K with rather long. pointe<l snout: snout rather large, with 
anterior canines stnmg, \ to jl: |KX<terior canines presi^nt: lateral teeth short, confluent in a serrated 
cutting edge; cheek and ojien-le si-aly: preo|K?rcle entirt*. l>oth limbs more or less si-aly; scales moder- 
ate, about 40 in lateral line: lateral line <*ontinuous: D. xii, 10; A. iii, 12; <lorsal spines low, pungent; 
soft dorsal and anal not elevate«l, their liai?es without si-ales: caudal subtrmiaite: j>e<*toral short. 

This genus is allie<l to Vernj and Sfj*ioU'.<. Frrun its nciirest relative, .VrxiYrf/Vf. it differs in the 
presence of a posterior canine t^Kjth. The single known si)ecies is l)rilliantly colore<l. 

36. Verriciilus sanguineus Jordan <fc Evermann, new species. 

Head 2.9 in length; depth 3.5; eye 6.2 in hearl; snout 3.1; mouth 2.8: interorbital 4.75; D. xii, 
10; A. in, 12; si-ale^ 5-40-13. 

Bo<ly el«»ngate, compress<;*fl oblong: hea<l lontr. pointe«l, conic, its depth 1.7 in its length; eye 
small, its posterior margin in middle of length of head: snout l«»ni:, |H;»inte<l, rounde<l; jaws produced, 
equal; mouth large, nearly horizontal, comer reaching l»i*low front rim of eye; lij>s thick, fleshy; teeth 
strong, thiise on sidt>« short, closi»-set, fonning a sharp cutting eilge on side of jaw; 5 i^anines in front 
of upper jaw, 4 in fn»nt «»f lower, a j>osterior <^nine on each side of upf>tT jaw; tongue long, pointed, 
free in front; preoj^eR'le not serrate: interorbital sj^a<v broad. c«:)nvex: iKjstrils small, anterior in short 
tube; dorsal spines strong, sharp-pointe<l, hmgest in middle and jjosteriorly; la-t dorsal si»ine 4 in head; 
anal spines ptn>ng, last spine longest, 3.75; seventh anal ray 3: i-audal n>un<le«l; don?al and anal fins 
scaled at base; pectoral roumled, 1.9 in heiul: ventrals short, spine strong, iK)inted, two-thirds longest 
ray, which is 2 in heail; caudal i)e<imicle bnxid, deep. 2.2 in head; sciiles small, thin, i-ycloid; head 
with verj' small tliin cycloid s<"ales on oi^cipnt, cheek, greater jjart of oi)ercle, l>ehind eye, and on 
opercles, otherwL»e nake«l; lateral line slightly curve^l in front, then obliquely down to base of caudal. 

Color in life, deep red, e<lge of upper jaw and lower tip golden; a long stripe lnm\ eye along l^iack 
to base of caudal golden, with a re<l shade, a vertical l>lack l>ar eilge<l with golden alx)ve, on opercular 
region; a long blai^'kish area covering it from eye toalx)ve |>ectoral, with some blackish before, behind 
and above; a black s}>ot at Ijase of caudal; dorsal and caudal golden, fir^t dorsal edged with violet and 
with the lower half violet; anal entirely dee]> bl«XHJ-red; ventrals golden; {sectoral retldish, golden at base. 

Color in akH.»hol, ver>' pale brown; a dusky Iwind from snout acn>ss back of head and on side, 
fading out indistinctly i>osteriorly; a blai'kish sfM)t at middle of base of caudal; opercle posteriorly 
with black vertical blotch; fins all j»ale or light brown. 


I)e8cril)e(l from the type, No. 5(X)77, IT. S. N. M. (field No. 0;W89), an example 7.5 inches long, 
taken at Hilo, with hook and line, in deep water with KUlim evuniHj EUimms marifhif Erythrichthys 
jt'hUujrily Antif/oitiit Hteindarhiwrt^ and AiUhiaa fiisr^ipinnia. 

We have examined oidy one example, the one descritei alx)ve. 

37. Pseudocheilinus evanidus Jonlan & Kvermann, new sptM'ieH. 

Head 3 in lenjrth; depth 3.8; eye 4.5 in hea<l; wiout 3; preorbital 6.2, interorbital 5.5; D. ix, 11; 
A. Ill, 9; sealeH 2-25-<). 

Iknly short, deep and oompreswHl; head lonjj, conic; snout lon^i;, 8hari)ly conic; anterior profile 
rising in a relatively straight line fn>m tip of unout to naiw, thence gently convex to haw of cau<l:il 
peduncle; ventral outline less convex; mouth large, horizontal, below axis of l>o<ly, fi:a|)e rt»ai*hing 
anterior line of orbit; upiHjrjaw wMth 3 pairs of antt»ri«>r canines, outer strongest, curvinl outwani 
and l)a(!kward; lower jaw with a single pair at tip, similar to inner alx)ve; jaws latenilly with a single 
series of smaller «mic teeth; preorbital narrow, oblique; eye high up, its lower lx>nlcr on axis of 
Ixxly; int<;rorbital si>ac« rather broati and flat; depth of caudal ixnlunde alnmt 2 in head; scales large, 
surfaces finely striate; hea<i, na|)e, and breast with large scales; lateral line following contour of 
l>ack until under base of sixth dorsal ray, where it is interruptetl, reapi>earing 2 rows farther down and 
continuing on 6 or 7 scales to liase of (.audal fin; fins rather large; dorsal s]>ines somewhat greater than 
eye in length, si)ines with a sheath of large scales reaching nearly to their ti{)s; soft dorsal and anal 
with a lower sheath; soft dorsal elevateil, rays injual to snout and eye; anal similar to soft dorsal, 
second spine strongest, nearly as long as snout; anal rays somewhat longer, e<|ualing those of soft 
dorsal; caudal rounded, its length 1.3 in head, its Imse covered with very largi*, thin wales. 

Color in life, according to Mr. Sindo, Ixxly dull brick-re<l; l)elly and l)astM)f anal pale purplish; 
about 17 thin, thread-like Kmgitudinal yellowish streaks along side anteriorly; dark greenish blotches 
alM)ve eye and on snout; a bhiish horizontal bar on cht^ek, Ik»1ow which is a yellow l)ar; nuMlian line 
of throat and tip of snout brick-red; edgt»s of oix'rcle and preoi:)ercle bright purple; a purple stri{>e 
with reildiah edges thnxigh middle of <lorsal fin, l)elow which the color is dull brick-reil, like that of 
bwly, and above which the spinous dorsal is orange-yellow, the margin t>f the membranes bright 
cardinal-red; alK)ve the purple streak in the soft <lorsal is a bright yellow strtak, alK)ve which the fin 
is cardinal-red, fading gradually upward; dorsal rays purplish; tip of soft dorsal somewhat rtnl; caudal 
rays purple, the membranes immediately next to the niys yellow, middle part <lull brick-red; anal 
same as caudal; ventrals pale purplish; pectoral \ya\v; iris sc^ir let-red. 

The same siK*ciinen after having lieen in spirits more than a yedr has the l)ody light bn>wnish- 
blue; a i)ale streak along each row of scales, but no trace of the narrow yellowish strtaks al>ove notecl; 
top of head and upper part of cheek dusky blue; oiH.*rcleand edge of i)reoiien?le rich blue; dorsid, anal, 
and caudal fins bright blue, the soft dorsal pale on ouU»r two-thirds, dorsal rays bright blue; ventrals and 
piHrtoral light blue, latter darker blue at base. The color of this s{>ecimen in spirits is wholly different 
from that which it ix)S8essed in life, and it would Im» «litficult to lK»li(»ve that such changes ha<l taken 
place except that the specimen was carefully taggei I in the field when the color note in life was taken. 

8inc<j writing the alx>ve, we have noticed similar changes in the Sanoan sjHH'ies, J*. hi\niiiriiui. 
The blue shades are jKjrmanent in spirits, while the pink or crimson wash s<K>n vanishes in spirits. 

The 17 thread-like streaks, mentioned in Mr. Sindo's field notes above, have vanishe<l entirely in 
the original type. A numlxT of sjHH'imens taken by Mr. Snyder at l^ysan, while c)n the AlfntfroM^ 
retain these traits, the strtaks Ix'ing almont white, like white tliR^ads, coverinir most of the side 
anteriorly. This is a very peculiar color mark, which should well distinguish the s|x»cies in life. 

A single specimen, tyi»e. No. 5067S, V. S. N. M. (field No. 05757), was taken by Mr. Sindo in 
Henshaw's pool near Hilo, a <lei»p tide-p(K>l in the lava nx!k8. 

38. Hemipteronotus baJdwini Jonlan A Kvermann, new spec'ies. 

Head 3.25 in length; depth 3; eye 5.75 in head; snout 1.75; maxillary 3; pnM)rbital 2.2; inter- 
orbital 4.8; I), ii-viii, 13; A. Ill, 13; scales IV-27-S). 

Bcxty m<Mlerat*»ly short and deep, grtatly c!ompn»sscd; head slightly dec»|>er than long; anterior 
profile marly vertical from mouth to inmi of eye, sharply cultrate; dorsal outline gently fonvex, sloping 
to the deep caudal iH3duncle; ventral outline less convex; caudal |)eduncle very narrow, the dei)th 
2.25 in head; mouth small, horizontal, the maxillary nearly naching vertical of orbit; the jawseciual. 


each provi<le<l anteriorly with a jxair of t^trong cur\'ed canines and laterally with a single row of short 
cloee-set conic teeth; lower jaw strong, ita outline very convex; preorbital nearly vertical and very 
deep; preopercle and opercle smooth, with membranoui? e<lgef:, the latter pro<luce<l «)mewhat in a 
broad roun<leil flap; origin of dorsal but little posterior to orbit, far in a<lvance ot Vmse of ventrals; 
first 2 dorsal spines somewhat remove<l from third but connei^te^l to it by a low membrane, their length 
scarcely greater than the gaf>e of mouth; remaining dorsal sjuned short and weak, warcely equaling 
gape; soft dorsal low, the rays slightly longer than the spines; anal similar to soft dorsal, rays some- 
what longer; caudal slightly convex, rays 2 in head; outer ray of each ventral somewhat prrnluceil, 
not reaching vent, the length alxjut 1.9 in head; pei-toral broad, the longest rays 1.7 in head; s<'aleH 
large, thin, smooth, firmly attached, those on breast somewhat reiiuce<i; heatl naked, except about 4 
series of email scales extending from eye downward to level of mouth; lateral line curving abruptly 
upward from upper end of gill-opening, following contour of Ixick to the si'ale undtT thinl dorsal ray 
from last where it drops 8 rows and continues to base of caudal, the iK>res simi)le, unbranche<l. 

Color in life (field No. 0312:^), pale, yellowish white over head and Ixxly; a diffuse lemon-yellow 
blotch under and alx)ve pectoral; a large brownish-black blotch on lateral line under seventh to tenth 
dorsal spines; dorsal tin yellowish-white, tip of detached jxart with a jet-black crescent { this marking 
variable in iKDsition, it sometimes l)eing farther posterior), rest of fin faintly mottleii with yellowish 
and olive, the latter in narrow oblicpie lines; caudal yellowish white; anal yellowish white, with 
narrow, waN-^-, j«ale-blue lines, and a large jet-black sjKDt l>ordere<l with blue on membrane of last 5 
rays; iris whitish. 

Color in alcohol, creamy yellowish white; head somewhat orange on cheek and opercles; faint 
rosy lines <lownward from eye to mouth and on preopercle; mediae line of anterior profile bluish; 
middle of back with a large black or brownish black blotch lying on lateral line, beneath which is a 
large white blot<h under and above pectoral fin ; anterior part of spinous dorsal blackish at edge, the 
color ocellate^l, rest of dorsal yellowish white with narrow purplish cross-lines; anal similar, with a 
large jet-black sjwt on last 4 ravs; caudal color of soft dorsal; jKK'toral and ventrals yellowish white. 

Color in alcohol, of one of the female cotypes (No. 03:572), pale olivaceous, the general I'olor that 
of the male; dorsal with ]>lack si>ots on membranes of siHond, thinl to fourth, and eighth spines, the 
latter ocellate<l ; a series of alx>ut a dozen small black si^^ts ]>ack of the dorsal blotch on side above 
lateral line; no black spot on anal. 

The al)Ove descrij)tion base<l uinm the tyi>e, No. 50644, U. S. N. M. (field No. 0:^14), a male 
example, 8.5 inches long, obtaineil at Honolulu. 

Another sjXH-imen, als<3 a male (field No. 0.*i*>71), was in life, livid gray; each scale pos-teriorly 
with a vertical sjM:»t of violet; anterior line of i>rofile bright vioU-t; a violet line downwanl from eye 
with a whitish art»a l)ehiud it on clu'ek; an oV>li(ine violet line downwanl and backward fn>m 
opercular fiap to l)ehind axil; lx*hind this a vague yellow area, l>ehind which is an ovate white spi>t, 
each scide around which has a vertical bar of bright violet; al)ove this a large black blotch washe<l 
with brick-re<l; dorsal bluish-gray, the rays jM)steric)rly with an increiisinjr amount of orange, where 
the blue is riMlucetl to ublicpie cnis.<l>ands, an intermarginal line of violet, a small black spot <in la^t 
ray; membranes of se<.*«»nd to fourth <lorsal spines with a terminal blark ocellus; anal j»ale golden, 
with oblitjue bluish stripes, a large jet-black «x*ellus IxjnieR'il with blue on Uist rays; caudal jjale 
orange, ctoss^mI by bluish lines; ventrals and jK*ctoral pale. 

Still another male example (field No. (^3004) was descril»e<l as follows: <ieneral color very |>ale 
smoky white, e<lgesof scales j)ale bluish, l)eneath seventh to ninth dorsal spines a large blotch, brick-re<l 
alx)ve, piile rosy l>elow, all irregularly overlaid with black or brown; mi^lian line from tip of snout 
to base of first dorsal spine bright ]»lue; a narn>w ])right blue line <lownwani from anterior part of eye 
to angle (A mouth; region alx)ve i>e<'toral j>ale lemon yellow, a short i>blique pale blue line'al>ove base 
of j^ectoral: dorsal i>ale flesh-color, with short vertical bluish lines, with 3 jet-l>lack siK>tsat tif«of first, 
second, and fourth spines; anal pale yellowish, a black sjxjt on distal half of last 3 rays; caudal i>ale, 
with ol>s<.'ure bluish cross-lines; pectoral and ventrals white; iris yellowish, red at lower i)osterior angle. 

Another examjjle, a female I field No. 0.*i372), 7.5 inches long, from ]I(molulu, which is taken as 
a cotype, <liffere<i in life coloration from the male in la<*king the black ocellus on the anal and in 
having more violet on the white lateral siM)t, also more gohlen l>efore it; violet lines and sp<.)ts oljficure, 
but present; 3 to 8 small blackish |xiints above lateral line Ix'hind black dorsal blotch; a small blac^k 
o<*ellus on second to thirfl dorsal spines an<l one on seventh dorsal spine, these wanting in some 
lemalet?; fins otherwise colored as in the males, but the blue fainter and the orange of dorsal brighter. 

F.C.B.19a2— 13 



Another female example (field No. 03005) differed in color from field No. 0iW)04 only in the 
al^sence of black on the dorsal and anal fins, the i»aler blue lineH on head, the paler caudal fin, and in 
having l^lack H|>otH on the ]>ack. Another female example (field No. 03271), 7.5 inches long, in life 
had the head and l)o<ly smoky white; a large bluinh white spot under tip of i>ectoral; snout bluish 
around lK)rder and surroundi^l by a bn)ad pale yellow spiice involving nearly all of anterior half of 
side l)elow level of eye; a large black spot under fifth to sixth dorsal spines, crossing lateral line and 
p<aietrating yellow of side, nearly reatrhing white spot; bac:k of this a series of alM)ut a <lo7A*n small 
black specks, scattere<l along side above lateral line to near end of dorsal fin; median line of snout 
and head ])lue; dorsal jiale, with wavy yellow cross-lines, pinkish toward margin; (^audal \y&\e; anal 
pale, with al>out a dozen pale yelU)w crossbars; pect<^>ral and ventrals pale; iris yellow an,d red. 

'This l)eautiful and abundant species is represent^^d in our collection by 41 siK'cimens, 19 of which 
are malefl and 22 females. The differences in coloration of the two sexes are very marked. The 
male, in all sjiecimens examineii, has the jet-black sjwt upon the last rays of the anal, a marking 
which is not present in any of the females examined. The female always has a series of small black 
8i>ecks on the side above lateral line posterior to the large lateral blotch. These markings, the small 
black spota on the side of the female and the large jet-l)Iack spot on the anal of the male, would 
apparently always serve to distinguish the two sexes. 

The extent of variation in color among indiviiluals of the same sex is indicatinl in the color 
descriptions given above. We should have added that occasionally there is a small jet-black 8[K)t 
ui)on the last rays of the dorsal. 

This is one of the most abundant and l>eautiful species found among the Hawaiian Islands. It 
appears to he related to i/. mdcgwptut of Hleeker, but differs from it markedly in the presence of the 
lai^ blai»k lateral blotch and in the absence of the large red lateral blotch shown in Bleeker's figure. 

This s|)ecies is named for Mr. Albertus H. Baldwin in recognition of his paintings of American 
and Hawaiian fishes. 



lA'ngth. Sex. 


8. 25 



Fem . 



Fem . 




Fem - 


Fem - 


Fem . 




Fem . 






do ... 





do — 














Final diHpt^wition of sfieoimen.s. 

Typo. No. 50644, U. S. N. M. 
i CotyiHj, No. 2713, U. S. F. C. 
I (^otype. No. '2714, U. 8. F. C. 

Coty[>e, No. 7162, L. S. Jr. Tniv. Mus. 

(VUype, No.3961. Field Col. Mub. 

(Vnype, No.3«;2, Field Col. Mu». 

Cotvpe, Nt>. 7462, L. S. Jr. Univ. Miw. 

Cotvpe, No. 60645, U. 8. N. M. 

Cotvpe, No. , M. C. Z. 

Cotype, No. , M. C. Z. 

Cx»tyi»e. No. 2290, -\m. Mus. Nat. Hist. 

Cotype, No. 2291 . Am. Mum. Nat. Hist. 

Cotvpe, No. 24219, Ac. Nat. S<'i. Phi la. 

Cotvpe. No. 24220, Ae. Nat. Si»i. Phila. 

(■<»tvi)e. No. 9809, Mum. Ind. Tniv. 

Cotype, No. 9H10. Mus. Ind. Univ. 

Cotype, No. 1495, Cal. Ae. Sci. 

Cotyi»e, No. 1496. Cal. Ac. Sci. 

C<»lvpt% No. 1696, Binhop Mu». 

Cotype. No. 1697, Biiibop Muk. 






Male ... 





do ... 









Male ... 






Male ... 


8.5 ... 



do ... 






do ... 



do ... 


7 ... 



— do ... 






do ... 


6.25 ... 


6.2 ... 


6.75 ... 


" 624 


Male ... 




a Collected by Jordan & Snyder. 

30. Xyrichthys niveilatiiB Jordan <fe Everinann, new species. 

Head 3.3 in length; depth 2.4; eye 6.2 in head; snout 1.8; preorbital 2; maxillary 3; interorbital 
4.7; 1>. ii-vii, 12; A. in, 12; scales 3-28-8. 

Body short, deep, and very greatly coniprt»88e<l; anterior profile nearly vertical from tip of up^KT 
jaw to front of eye, thence in a jmrabolit; curve to dorsal fin; anterior dorsal outline very trenchant; 
body tapering rather evenly from head to caudal i)eduncle, which is gn*atly compn^ssed and very 
deep, depth at middle eijualing preorbital; head short; snout very short and blunt; mouth small, hor- 
izontal, the maxillary nearly reaching anterior edge of orbit; jaws equal, each with a }»air of strong 
curved canines in front, and a single series of smaller, conic teeth laterally, the canines of lower jaw 
most prominent and extending in front of ujuHjr jaw; eye small, high up; the interorbital spatv narrow 
and trenchant; opercles smooth, without spines or serrations, ending in thin flexible edges; preorbital ^ 



vertical and very deep; origin of dorsal fin above iKX?torior line of orbit, far in advance of \fSL8e of ven- 
trale; first 2 dorsal jspinen Homewbat roniove<l but not detache<l from tbird, tbe nienil)rane l)etwi'«'n 
second and third spines ni<xlerateiy notched, length of neeond Hpine alxmt 2.7 in head, n*imiininjir dorsal 
spines Hube<iual, weak, a^K>ut equal to gap<»; dorsal rays low, the la^-t few somewhat pnMluctMl, their 
length 3 in head; anal similar to softdorKd; caudal short, slightly convex, rays al)out e<iual to preorl>- 
ital; outer ray of ventral somewhat produceil, not n*aching vent, its length e<iualing depth of pn»orbital; 
pectoral broa<l, its tip reaching vent, its length (^jualing <listance from snout to e<lge of pn*oix*rcle. 

Scales large, thin, and with meml)ranous e<lges, those on breast somewhat smaller; head entirely 
naked, ext*ept for a few small scales Inflow the eye; lateral line l)t»ginning at upper en<l of gill-ojjening 
following closely the curvature of back to the scale under the la^^t dorsal ray but 2, when* it drops 3 
scales and continues to ba«e of caudal, the pores simple, rarely l)ranche<l. 

Color in life, grayish; each scale of iM>sterior half of IkmIv with a large violet spot, more narrow 
and brighter near middle of Ixxly, the edg<^ of each scale broadly goMen-olive; a large golden ana, 
anteriorly deep orange, alx)ve pectoral and on e<ige of open'le; Ix'hind this a large <iuadrate i)ure white 
area extending to tip of pectoral; a few scales in golden an*a with bright violet markings; hea<l sha<leti 
with violet, a bright violet stripe downward from eye to behind angle of mouth; a lunate black area 
sliaded with red just l)elow front of soft dorsal; spinous dorsal violet-gray, edgeii with reildish; soft 
dorsal golden, with violet vermiculations at base, its edge orange; anal golden, with bluish vermicula- 
tions; caudal similar, with the bluish markings; |x»i'toral faintly re<ldish; ventrals dirty white. 

One of the cotypes, a male (field No. 03373), agreeil in life coloration with the type except that 
behind the o[)ercle is a golden area with the bright violet stripes a<Toes anterior basal part; behind 
this a large milk-white patch beyond tip of i>ectoral; a violet Iwrder was around the white and 
blackish above the vellow. 

Color in alcohol, dirty yellowish white, dusky alx)ve; head with some purplish n»flections; a thin 
purplish line downward fn>m anterior edge of orbit to tip of maxillary*, a similar but less distinct line 
from humeral region downward to subojierde; a yellowish white blotch on side above base of jH^ctoral, 
in the base of which are 2 or 3 small purpli.^^h spots; a large white area on middle of si<le under and 
above tip of pectoral, sej>arated fn)m the yellowish blotch by purplish brown on 2 or 3 scales; a black 
siM>t covering the larger part of 3 scales on side al)ove lateral line under l>ase of first 3 dorsal rays, l>ack 
at base of last dorsal rays somewhat dusky; anterior j>ortion of dorsid fin dusky olivaceous, soft dorsal, 
anal and caudal pale yellowish crossed by narrow, wavy, pale purplish lines; ventrals and pectoral 
plain yellowish white. 

This handsome fish is rather common al)out Honolulu. 







Or>164 . 

y. 7;-> 


•.». 2:> 


H. ■_♦ 




*-. •"» 


Finul di>iM>sition of }*[MTinu'n. 

Honolulu Tv|K'. No..'><Nl4fi, r.S. X. M. 

do Ci.tviH', N«».:jyti<), Fk-ld Cnl. .Mu.«i. 

do CotyiK'. No. 271'>. W S. F. ('. 

do C^oiyiK.', N<». l\(Kt, L. S, Jr. rniv.Mu*. 

do ' roty[K\ No. , M. C. Z. 

do Cotypt*. No. 22y2, Am. Mu-s Nat. Hist. 

Family SC.\KID.4:. The Parrot- fishes. 
40. Scarus jenkinsi Jonlan <& Evermann, new s]>ecies. 

Head 3 in length; depth 2.5; eye 6.5 in head; snout 2.6; preorbital 4.7; intemrbital 3; I), ix, 10; 
A. Ill, 9; P. 13; scales 2-24-7. 

BckIv short, very deep ancl greatly compressed; head short, nearly as det^p as long, snout short 
and blunt; mouth small; (?ach jaw with 1 or 2 blunt canines; dorsal and ventral cjutlines alniut ecjually 
convex; anterior profile rising rather irregularly from tip of snout to origin of dorsal; caudal i>eduncle 
deep, its least depth 2 in head. Scaler large, deei)er than long; 2 rows of largt* scales on cheek and 1 
row on subopercle; a row of thin mcnlified scales at base of dorsal and anal; a few very large, thin 
scales on base of caudal; lateral line ceasing under last dorsal ray, nap[)(*aring 2 rows lower down and 
continuing to base of caudal, the pores with 2 or 3 irn»gidar branches; dorsal rays soft and flexible, 
not pungent; dorsal spines somewhat elevate<l inisteriorly, longest a little mon* than 2 in head; first 
ventral spine obscure, the others scjft and flexible; anal rays somewhat shorter than those of dorsal; 


caudal shallowly lunate, the outer rays not preatly pro<luee<l; ventrals moderate, 1.6 in head, not 
reaching to origin of anal by a distance equal to two-fifths their length; i)ectoral broad, 1.2 in head. 

Color of a nearly fresh specimen, bright blue-green, brightest on posterior half of Ixxiy, each 
scale broadly edged with reddish brown; lower anterior part of Ixxly reddish brown, with traces of 
blue-green; top ()f head brownish red or (;oppery, a broad deep blue-green band on the ui)|)er lip, 
extending on side of head to below eye; lower lip with a narrow brighter blue-green Imnd connecting 
at angle of mouth with the one from upi)er lip; chin with a broad coppery-rtni bar, followed by a 
broader bright blue-green one; caudal green, meilian part pale, banded with greeh spots; dorsal 
bright green at base and tip, the middle pale greenish, translucent; anal similar, the distal band 
broader; pectorals and ventrals deep vitriol-green with whitish markings. 

Color in alcohol, dirty greenish, side with about 8 longitudinal series of greenish blotches; head 
olivaceous above, paler on cheeks; upper lip broadly i^ea-green at edge, this color continueil l)ackward 
to under eye; edge of lower lip pale green, continue<l around angle of mouth uniting with the same 
color from upper lip; chin with a broad, pale crosslmr behind which is a broader, pale green one 
which extends up on cheek nearly to orbit; back of this is a still broader, white crossbar interrupted 
in the middle by greenish; subopercle and lower edge of preopercle with a large, irregular, green patch; 
a median green line on breast to base of ventrals; dorsal green at base and along edge, the middle 
portion paler; anal similar to dorsal, the green border broader; caudal bright pea-green on the outer 
rays, the inner ones pale with 4 or 5 cross-seric« of green spots, tips of rays darker; ventrals pale 
green, the edges dark pea-green; pectoral pale greeu, darker grei*n on the upper rays. 

This species is related to Scarus giWerti from which it differs in the greater depth and the somewhat 
different coloration. It is also related to Scarus lanioy but differs in the much greater depth, the 
lees produced caudal lobes, the greater width of the green head markings, and the color of the fins. 

Only one si)ecimen was obtainwl, type, No. 50647, U. S. N. M. (field No. 02944), 14 inches long, 
obtained at Honolulu, June 6. Named for Dr. Oliver P. Jenkins. 

41. Scarus lauia Jordan & Evermann, now species. '^Lauia,'* 

Head 2.8 in length; depth 2.7; eye 6.75 in head; snout 2.6; preorbital 4.8; interorbital 2.8; D. ix, 
10; A. HI, 9; P. 13 on one side, 14 on other; scales 2-25-6. 

Body short, stout and compresseii; head heavy; snout rather short, bluntly rounded; dorsal and 
ventral outlines a}>out equally arched, anterior profile slightly concave befort* the eyes; najx^ strongly 
convex; mouth small, nearly horizontal, in axis of Ixnly; upp<»r jaw with 1 or 2 moilerately strong, 
backwardly directeil canines; a similar but smaller canine 8ometimt»s present on lower jaw; cutting 
edge of upper jaw fitting outside that of lower; teeth white; eye small, entirely alx)ve axis of Ixxiy; 
opercle with a broad short flap. Scales large, their surface with line lines and granulations; nape and 
breast with large sc^alee; cheek with 2 rows of large scales, alx)ut 7 scales in each; sul)opercle and 
lower limb of preopen-le e.ach with a row of scales; opercle with large scjiU*s; lateral line broken under 
last dorsal ray, reappearing one row lower down and continuuig to caudal fin, the pores with 2 to 4 
branches; a series of thest^ oblong scales along l)ase of dorsal and anal; bas<» of cau<ial with 3 (►r 4 very 
long, thin scales. Dorsal spines soft and flexible, not pungent, the longest al)out 2.7 in hea<l; soft 
portion of dorsal somewhat higher, es})ecially iM)steriorly where the rays an* alKuit 2.4 in head; anal 
spines soft and flexible, the first obsirure, the third al)out 4.3 in head; anal rays higher, the last but one 
longest, 3 in hea^l; caudal deeply lunate, the 3 or 4 outer rays above an<l lx»low pnKluceti, length of 
middle rays 2.3 in head, or 2 in outer rays; ventrals mo<lerate, not reaching vent, 1.9 in head; pectoral 
broad, the free margin oblique, length of longest rays 1.3 in heii<l. 

C/olor in life, head brownish yellow before eyes, the jaws lighter yellow; cheek washed with ])rown- 
iflh and blue, throat greenish; nuchal and oi)ercular regions brownish orange; Inxly salmon-color 
' above, the l)elly lighter yellow, most of the scales with an edging of greenish blue; a deep blue line 

from nostril Ixifore and behind upptT part of eye; upper lip deep hhw, the streak forming an inter- 
rupted line before eye; lower jaw with 2 blue cross-lines, 1 marginal; a dark blue spot l)ehin<l angle 
of mouth; deep blue blotches on interopercle; dorsal deep blue with a f)eculiar jagged stri|>e of light 
brownish yellow; anal with blue spots at l>ase, then light yellow, then deep blue, then green with blue 
edge; caudal brownish yellow, with bright blue edgings and a me<lian area of bright golden green; 
ventrals golden, trimmeil with bright blue; pectoral golden with deep blue above and greenish biye 
on lower rays, a salmon streak across base with gri^nish blue behind it 

Color in life of another example (No. 0,3040, 10 inches long), i>ale coppery rosy, darker on first 3 
rows of scales; the center of each scale in the first 5 rows greenish blue; under parts pale rosy, with 



orange wash; head i>ale rosy, a small postooular blue spot, a short blue line forwanl from eye, and a 
eeft.>ud of same color on upper lip and across cheek to eye, where it ha^ a slight break, then continues 
under eye as a greenish-blue bar; under lip with narrow ]>lue inlge; chin fa«le<^l salmon with a double 
blue crescent; sjiace from chin to isthmus bright blue; an oblong bright ]>lue 8iH)t on suboj^ercle, 
behind which is a smaller irregular one b<>nlere<l above by a l>roa<l greenish-bhie space; tlorsal green- 
ish blue, with a broad submedian orange l>antl, the lower greenish-])lue band made up of large, s<arcely 
connecte<l, l>luish spots, the upper half c<>ntinuous with a narrow bright blue bonier; a small orange 
Idotch on base of last dorsal ray; caudal i»ale rosy at base, then with a greenish Iwir, folluweil by a 
broad n>sy bar, then by a broad teniiinal greenish-bhie bar, dark blue in front, greenish in middle 
and pale blue on outer third; uppi-r and lower e<lges of caudal blue, below which is a broa<l rosy 
orange strij^; anal greenish blue at ba.<<\ then a broa<l orange strij:>e, the ruiter half greenish blue with 
narrow bright blue edge; pectoral orange anteriorly, j^ale bluish I jehind, the anterior liorder blue; ven- 
trals orange, anterior e<lge and tip blue: iris j)ale orange. 

Color in spirits, light dirty grayish white, lighter Ijelow; a narrow pea-green strij)e on edge of 
upper lip, breaking up into irregular si>ots fr«»m angle of mouth to lower e<lge of orbit, a similar stripe 
from nostril to eye and slightly l)eyond upi>er iK)sterior l>on.ler of eye; these lines sometimei* continu- 
ous and unbroken; lower jaw etlgeil with green, a broader i)ea-green cross-strijn^ at anterior edge of 
branchiostegal oiK?ning; suboi)ercles each with a broad green strijx?; line of union of gill-membranes 
broadly green; dorsal with a series of large olive-green sjx)ts at base and a bnuid l>an<l of similar color 
on distal half, these separateil by a paler l>and and cut by intrusions from it }^)Xh al>ove and l^low; 
dorsal fin with a very narrow paler bonier; and with a serii's of greenish sj^ots at base, then a broa<l 
pale yellowish white line, boundeil distally by an indefinite, wavy, black line shading off into the 
greenish of the distal half; last ray of anal dusky on its outer thinl; caudal greenish-olive at lyase and 
on produced outer rays, eilges of tin al)Ove and l>elow green; middle rays with a brr)a<.l lunate area of 
pale green, 8calloi)e<l proximally by dark green, separate<l from the lighter green base by a broad 
whitish interspace, the upj>er and lower edge also darker green; ventrals creamy white, the outer edge 
pale greenish; j)ectoral whitish, the uppi'r e<lge dusky. 

There is some variation in the width of the green markings on the head, sometimes the stripes on 
the lower part of the head lx»ing verj* broad. 

This species is relate^l to Styirns gilfn-rti, from which it differs in the more strongly pnnhiceil caudal 
lobes, in the narrower lines on the snout, the broader green lines r>n the throat, the al>sence of a green 
me<!Kan line on the breast, and in tne ver\' different coloration of the this. 








! do 

I <lo 

: do 

Final dispof^itlon of .*jj>eoimen. 

,' Type. No. 50t>48. r. P. N. M. 

, I Cotype, No. 74M. L. S. Jr. I'niv. Mus. 

.' rotyiK-. No. 2711). r. S. F. C. 

. Colvpe, No.ysil. Ind. Tniv. Mu«. 

. , Cotype, No. 3964, Field Col. Mas. 

Aljo one specimen i^No. 12046). 9 inches long, o))talne<l by Dr. Wotxl at Honolulu. 

42. Scarus barborus Jonlan «fe Evermann, new species. 

Hea<l 3.2 in length; <lepth 3.2; eye 6.6 in head; snout 2.9; interorbital 2.9; preorbital 4; D. ix, 
10; A. Ill, 9; P. 14; scales 2^2.>-6. 

l^nly oblong, not very deep nor greatly compressed; hea<l about as long as deep, conic, com- 
pressed; snout short, blunt and n)unde<l; upper jaw proiluced, its lip double, C4>vering entire dental 
plate; lower lip covering half of <lental plate; no canine teeth; eye anterior, high, itrf lower bonier con- 
siderably alx>ve upper ba^^e of ])ectoral; caudal jXHluncle short and deep, its depth 2 in hejid. 

Origin of dorsal over up]>er ba.»^* of j^ctoral, spines flexible, short, not quite as long as rays; long- 
est ray 2.1 in head; lt»ngest anal ray 2.2 in head; caudal truncate; ventrals 1.9 in hea<l, not reaching 
vent by half their length; ixrtoral 1.5 iu head. Sc:des large and thin, very slightly roughened by 
radiating lines of granulations extending to margins of s<.*ales; lateral line intemipted, the i)ores being 
on 18 scales, then dropping 2 rows to row of s<^les under i>osterior base of dorsal, and continuing to base 
of caudal on middle of caudal iHHluncle, 7 j»ores in the shorter part, which l)egins on the row following 
the row on which the upper part ends, there not l>eing 2 pores in the same njw; the scales extend well 



; i 

I < 
■ t 


. t 

out on the caudal, the last scale of lateral line, very large and thin, being the largest scale on the fish; 
4 scales in median line l)efore dorsal; 2 rows of scales on cheek, 5 scales in upper row and 2 to 4 in 
lower, sometimes only 2 on posterior part; 2 rows on opercle, and 1 on lower margin. 

Color in alcohol, grayish leaden brown, lighter below; no markings on fins different from corre- 
sponding parts of body evi<lent. 

The above description is based on the type, No. 50649, U. S. N. M. (field No. 04316), a specimen 
7.75 inches long, from Honolulu; cotype, No. 27J^, IT, 8. F. C. (field No. Q^i'H), 7.5 inches long, from 
Honolulu; cotype. No. 7465, L. 8. Jr. Univ. Mus. (field No. 650), 5.5 inches long, from Honolulu. 

Family TEUTHIDIDiC. The Tangs. 
43. Teuthis atrimentatus Jordan & Evermann, new species. 

Head 3.8 in length; depth 1.9; eye 4.2 in head; snout 1.2; interorbital 3; D. ix, 27; A. in, 25. 

Bo<ly deep, compressed, ovoid, the upi)er profile steeper than lower, evenly convex; jaws low, not 
produced, lower inferior; mouth small, inferior; t^^eth broatl, compressed, e<lges crenulate; nostrils 
close together, anterior larger, with small fleshy flap; anterior dorsal spines graduated to posterior, the 
longest 1.5 in head; fourth dorsal ray 1.4; third anal spine longest, 1.9; first anal ray 1.5; caudal large, 
emarginate, upper and lower rays produced in sharj^ angular points, upper much longer than lower; 
I)ectoral about 3.5 in body; ventrals shari)-pointed, 3.6 in body, spine half the length of fin; caudal 
I>e<luncle compressed, 2 in head; caudal spine large, deprcssible in a groove, 3.1 in head; scales very 
small, ctenoid, few, and very minute on vertical fins; lateral line high, archeil, at first descending 
un<ler fifth dorsal spine, then straight to below middle of soft dorsal, finally falling down and running 
along side of caudal i>eduncle to tail. 

O>lor in life (No. 02996), coppery brown, crossed by numerous, very narrow, pale blue lines, those 
above axis of Inxly ninning somewhat upward and backward, and with short broken lines of same 
interspersed, those l^elow more Regular but less distinct; cheek brassy, with alx)ut 5 narrow pale blue 
lines from eye to snout; a conspicuous jet-black spot on caudal peduncle at base of last dorsal ray, eai^h 
of these* extending slightly upon pale rusty, each with 5 or 6 narrow brassy lines parallel with margin; 
e<lge of each blackish; last niys of dorsal and anal more brassy; caudal dark, blackest on outer part of 
middle rays; pectoral pale lemon; ventrals dusky, blacker toward tips; iris brownish, white on 
posterior part. Another example (No. 03474) was dull olive-gr^, unmarked, save a faint whitish 
band across nape and l>ack part of head ; fins plain dusky gray. 

C<jior in alcohol, very dark chocolate brown; side with about 40 narrow irregular or incomplete 
series of indistinct dark slaty longitudinal lines; cheek with similarly colored lines running obliquely 
downward; fins, except pectoral, all more or less blackish or dusky; dorsal with about 5 blackish 
longitudinal bands; anal with several similar indistinct blackish bands; base and axils of last dorsal 
and anal rays blackish ; i>ectoral brown. 

This common Hi>ecies is well distinguished from Teuthis diistnimieri and other streaked species by 
the black ink-like sjwt in the axil of the dorsal and anal fin. It has several times been recorded 
under the erroneous name of Acanthunis lineohttitSy but the species originally called by that name must 
1)0 something else. Numerous specimens were obtaincnl by us at Honolulu, where it was also secured 
by Dr. Jenkins and Dr. Wood. We have examined the following examples: 


: I 


I ' 







do ... 


5. 75 

d<) ... 

do . . . 

do . . . 


6. 75 

do ... 



do ... 


5. 2 

do ... 



do - . . 

Final dLsiRMiitiou of 8p<*ci- 

Typf, No. 50673, U.S. N. M. 

Coty^, No. 7482, L. S. Jr. 
Univ. Mus. 

CTotypo, No. 2731, U. S. F.C. 

(•^)tv|H', No. , M. (\ Z. 

. CotviH?, No. 2301, Am. Mus. 

I Nat. Hist. 

i Cotvpe, No. 24229, Ac. Nul. 

I Sci. I'hila. 

' (^otyK, No. 39<k5, Meld Col. 


Cotype, No. 9818, Ind. Univ. 


('otyjH.', No. 1707, Bishop Mus. 














5. 75 
















O. P J. 








Final dLsponition of Hpoci- 

Cotype, No. 1501, Cal. Ae.Scl. 








Cot y pi'. 

Adinthuruif lintxdiUuit, (iihitber, Fisehe der Hiidsee, i, 112, taf. i.xxiii. fig. A, 187:^ (Sot'iety Islauda); Stelndachner, l>enk.s 
Ak. Wlas. Wlen, lx.x, 1900, 493 (Honolulu); not of Cuvier tt Valenciennes. 



Family BALISTID.^. The Trigger-6shes. 

44. Pachynathus nycteris Ionian A Evermann, new gpeciee. 

Head 3.5 in length; depth 1.9; eye5inht'a'l: snout 1.25; interorbital 2.6; preorbital 1.5. D. iii-33; 
A. 29; scales' aV»out SO. Bixi y short, stout, deep, and greatly roniprease<l; lu^ short; <lon»l and ventral 
I»rotilej* alx>ut equalh*eurve<l; <*audal j:»i'<luncle short, compressetl, its leatiit depth al>out twi« v diameter of 
eye, its least width aV>out e<}ual to diameter of eye; a short horizontal gnxne in front of eye l»elow no:*trils; 
nostrils small, close t4)gether in front of upi»er part of eye; teeth brciail, <'lose-set, forming a continuous 
plate, the teeth, however, not nnitiHl; lii»sthin; mouth small, horizontal, in axis of Ixxly; lower jaw very 
slightly the longer; gill-oj^ening short, nearly vertical ; a group of In^ny .s<*utes under i>ectoral liack of gill- 
opening, one of these c^onsiderably enlarge<l; s^-ales regularly arrange«l in rows, their surfaces granular; 
lateral line beginning at posterior e<lce of eye, ascentling to within 7 scale:? of ffjiinous dorsal an<l ct>n- 
tinuing to near origin of soft dorsal, where it disappears; scales on i>osterior portion of iKxly an<l on 
cau<lal pe<luncle eat^h with a slitrhtly raise<l cref^t at its cent4.T, these fonning series ai ritlges along the 
side. First dorsal spine St nmg, l)lunt, ami rough, its length alnuit 2 in hea<l; second dorsal spine 
shorter and much weaker, iu* length scarcely more than one-thim that of first; thin! dorsal spine 
remote fnjm the second and very short, not extending alcove <lorsal groove; 8i)ft dor^sil gently rounde<l, 
its rays of approximately eipial length, the longest e<jualing the distance from tip of snout to posterior 
edge of eye; base of soft dorsal slightly greater than distance from tip of snout to i>08terior base of first 
dor>7al spine oretiualing distance from tip of snout to Inwer Ijase of jjectoral axil; anal similar t4) soft 
dorsali the rays s^^mewhat longer than thr>se of s<jft dor«il, the liase somewhat shorter; cau<lal short 
and rounded, the rays alx>ut 1.75 in hea^l; pet-toral short, the upper rays longest, ab<»ut 3 in Iiea^l. 

CJolor in alcohol, rich l>rownish or velvety l>Iack; spinous tlorsal black; soft dorsal j»ale yellowish 
Dr whitish, margined with black, the lower half crost^e*! by 4 narn»w (larallel black lines; anal similar 
to soft dorsal, but with only 2 narrow black linea on its basal half; caudal dusky, yellowish at tip; 
pectoral yellowish. 

Only one specimen. Type, No. 50821, U. S. N. M. ^^ field No. 05089), 0.25 inches long, Honolulu. 

Family TETRODONTID.^. The Puffers. 

45. Ijagt>cephalus oceanicus Jonlan (.V Evennann, new spe<'ies. 

Head 2.8 in length; depth 3.6; eye 4.5 in head; snout 2.4; interorbital 3.2; depth of caudal 
peduncle 6; D. 12; A. 12; C. 10; P. 14. 

Bo<ly rather el«»ngate, mo4lerately compressed, greatest depth at vertit^l of pe<"toral; head long; 
snout long, blunt at tip, the sides riattene«l; anterior profiU* fn:»m tip nf snout to vertical of pectoral in 
a long, low, even curve; ventral outline little convex when not infiate<l; mouth small; ttvth iK»inte<l at 
median line, the cutting e<lge shari>; n<»strils sejiarate, nnt in tul>es, the anterior somewhat the larger, 
their di.*»tance from eye about lialf their di.'^tance from snout or al>out half the interor]>ital sp;ict*; gill- 
opening vertical, 1.2 in eye, extending a little alxne l»ase of ]»ect<»nil, inner flap entirely hidden by 
outer; eye rather laiige, wholly alxjve axis of Ixxly; inten)rl)ital si>ii<*e very little convex; cheek U)ng; 
caudal jxHiuncle nearly round, tapering, it< length fn>m anal fin eijualing snout; liack, upj>er parts < f 
sides and entire ht»ad entirely smo<>th, no spines or prickles evident; l»t*lly covere«l with small 4-n:M>ted 
spines, most prominent when lx»lly is inflate<l, spiniferous area not extending on thmat anterior t«j eye, 
nor on side al)ove l>a.«e of ptvtoral. but in fn^nt of anal extending upward to level of lateral fold; a line 
of very small mucous jn^res curving alx>ve eye on interorbital .•^>a4-e; a strong cutanec>us fold on lower 
part of side <jf caudal fKxluncle (nmi alx)ve anterior l>ase of anal to lower l» of cau<lal fin; no denual 
fold on head or anterior part nf ImxIv; mucous |H)res inct^nspicuous; dorsal fin somewhat anterior t<» 
anal, pointed, anterior rays prtMluce*!, their length c>qual to that of .snout; anal similar to dorsal, its 
rays somewhat longer; (*au<lal lunate, outer rays about 2 in heail; jxvtoral l»road, its length a little 
greater than .snout, 2.3 in head. 

Color in life, back blackish, fading into deep steel-blue on side; side and lx?low from level of 
upper e^lge of eye abruptly silvery-blue; sides of l>elly white, with roun«l black sjMits about as lai^ge as 
pupil, these must distinct about pectoral, liefore, lielow, and Ixfhind the fin; upper fins dusky; caudal 


1 ! 
i ■ 


mottled black, tipped with white; |>ertora] black above and behind, pale l)elow; anal |>ale, broadly 
tii)i)e<l with blackish. 

(■olor in alcohol, bliiiph black al)ovo; fii<le from upper level of eye abruptly bluiflh nilverj'; back 
crossi^d by 7 or 8 narrow <larker «To88-streaks; lx»lly white, with a nerieH of aU)ut 9 to 12 small 
roundish black siK)ts, chiefly l>elow the j)ectonil; cheek dusky; i>ectoral, dorsal, and caudal dusky, 
tii»s of the latter paler; anal whitish, a little dusky at tip. A somewhat smaller example (4.5 incheH 
lon^') has larjjer <lark spots alon^ middle of sidti alK)ve level of pectoral. 

This siM'cics is known to us from 2 small examples obtainwl in the market of Honolulu. It is 
related io Ijujocephahin mteilntu^ (Donovan) of Europe (Tetrodov hiywrphnlus of (iiinther, not of 
Linmi'us), but differs in the much shorter i)e<'t oral, more conspicuous siK>ts, and rather greater extension 
of the ])rickly region of the breast. Thetyi>es of TetrmUm hujo<rphdm Linuicus are repute<l to have 
come from India. According to Linmeus, this siHJcies had 10 dorsal and S anal rays. It may have 
l)een bast'd on hi(jorf:ph(ilu>t sreUratmt or sonu* other East Indian sjH»cit»s, but there seems to l)e no 
evidence that it was identical with the Euroix'an L<i(joc.ephalu» mtcllatuK. In any t*vent, the Hawaiian 
fonn seems dafferent from anv other vet known. 

TyjM?, No. 50820, V. S. N. M. (Oeld No. 03871^), 5 inches long, obtaiuinl at Honolulu; cotyixs No. 
7784, L. S. Jr. Univ. Mus. ( field ^o. 534, pai>er tag), 4.5 in<*hes long, also from Honolulu. 

Family OSTRACIID.'E. The Trunk-fishes. 
46. Ostracion oahuensis Jonhm A Evermann, new species. *^Momo Auna.^^ 

Head 3.9 in length; depth 2.9; eye 2.9 in head; snout 1.2; pRH)rbital 1.0; interorbital 1; D. 9; 
A. 9; P. 10; C. 10. 

Body 4-sided; dorsal side of carapace evenly cx)n vex, its greatest width one-fourth greater than 
head; lateral dorsal angles not trenchant, slightly convex anteriorly, then evenly c(»nvex; snout blunt, 
the anterior profile as<*ending abruptly then strongly convex in front of €*yes; interorbital space nearly 
flat; cheek flat; side of lM>dy concave, its width al>out iMpial to head; ventral keel prominent, evenl\ 
(ronvex; ventral surface nearly flat posteriorly, but little convex anteriorly, its greateijt wi<lth 1.4 tim«' 
h^ngth of head, its length just twice its width; gill-o])ening short, not exceeding two-thirds diamet*i 
of eye; lea*<t width of anterior opening of <'arapa<-e 1.75 in interorbital, or 1.5 times diameter of orbit, 
the depth nearly twice orbit; mouth small; tei»th rich brown; least dejith of posterior o|>ening of 
cjirapiK-e much less than width of anterior o|)ening, equaling distance from lower e<lge of preorbital to 
jnijiil; length of cAudal j>eduncle less than that of head, its depth 2.2 in it** length; no spines anywhere. 
Dorsid fin high, its e<lge obliquely rounded, it*< length 1.3 in head; anal similar to dorsal, the edge 
rounde<l, it« length 1.2 in dorsal; caudal slightly rounded, its rays nearly eijual to head; pectoral with 
its free e<lge obliijue, the rays successively shorter, length of fin e<iual to height of dorsal. 

Color m life, dark brown with blue tinges; interorbital space showing more or less golden; small 
whitish spots pn.)fusely covering entire^ dorsal surface?; no siK)ts on side of bo<ly or on face; no s]X)tsou 
ventral surfac-e except a faint one of a slightly darker color than general gray color of surface; one 
longitudinal row of golden spots on each side of upi)er i^art of caudal j>eduncle from carapac^e to base 
of ciiudal fin; pectoral, anal, and dorsal fins with transverse rows of faint spots; caudal bluish black at 
base, white on posterior half; a broad light or yellowish sltva Ixdow eye; iris golden. 

Color in alcohol, rich bmwn above, the sides darker, and the ventral surfa<*e paler, brownish 
alwut margins, dusky yellowish within; entire back with numerous small, romidish, bluish-white spot*; 
ui)per half of <*au<lal i>eduncle with similar but larger si^ts; forehea<l and snout dark bnjwn; lips 
brownish black; chwk dirty yellowish; sides and ventral surface wholly unspotte<l; base of caudal 
bla<rkish, ]>aler distally, the <lark extending farthest on outer rays; other fins dusky, with some ol)8t!ure 

•brownish siK)t*«. 

This si>e(!ies is relattxi to O. mm?imm Jenkins, from which it <liffers in the smaller, mon^ numerous 
spots on back, the entire absence of six)t? on side, the smaller size of the spot8 on the ciiudal ptvlunde, 
and the brighter yellow of the sulK>rbital region. Only 2 gpwimens known, ]»oth from Honolulu. • 

Type, No. 50()<>8, U. S. N. M. (field No. 03443), a sptnimen 5.<> inches long, obtaine<l by us at 
Honolulu, July 25, 1901. CotyiK?, No. 7478, L. 8. Jr. Univ. Mus. (field No. 2166), an example 5.25 
inches long, collected at Honolulu, in 1898, by Dr. Wood. 


Family SCORPi€NIDi«. The Rockfishes. 

47. Pterois sphex Jordan & Evermann, new species. 

Head 2.4 in length; depth 2.f>5; eye 3.8 in head; snont 3.2; interorbital 5.2; maxillarj' 2 S5; 
mandible 2; I), xiii, 11; A. in, 7; V. 16; V. i, 5; scales 10-56^13. 

Body elongate, coinpreased, greatest depth at first dorsal spines; bat^'k only slightly ele\Aced; 
snout rather short, roundinl; mouth large, oblique; maxillar>' reaching lx»low anterior rim of orbit, 
its distal expanded extremity 1.75 in eye; teeth fine, in bauds in jaws and on vomer; lips rather 
thin, fleshy; tongue jxiintetl, compressed an<l free in front; jaws nearly equal; eighth dorsal spine 
longest, equal to head; penultimate spine 4; fifth dorsal ray 1.75; third anal 8j)ine longest, 2.2; third 
anal ray longest, 1.5; caudal rounded, elongate, 1.4; pectoral long, the rays more or less free for 
at least half their length; ventral 1.3 in head, reaching l>ey(md origin of anal; ventral spine 2.1; 
caudal peduncle compressed, its depth 3.75; nasal spine very small; preocular, supra<Kular and post- 
ocular spines present, the upper lx)ny ridge over eye l)eing serrate; tympanic, coronal, parietal, 
and nuchal spines present, coronal ver>' small and close together and parietal with 4 serrations; a 
finely serrate ridge from behind eye over opercle to suj^rascapula; a finely serrated ridge over 
preorbital and cheek to margin of preopercle, ending in a strong spine, l)elow this 2 other spin^; 
preorbital with a strong spine over maxillary i)Osteriorly, and with fine serrations above; scales 
ctenoid, present on top of head, cheeks, and oj)ercles, head otherwise nakiKi; tubes of lateral line 
single, in straight line to base of caudal; several fleshy flaps on head, 1 al)ove eye, 1 from lower 
preorbital spine, and 2 from along margin of j>reopen*le. 

Color in alcohol, very pale brown, whitish l)eneath; side with 9 broa^l, deep brown bands alternat- 
ing with narrow brown bands on trunk and i>osterior portion of head ; narrow brown l)ar8 from below 
penultimate dorsal spine with a narrower brow^n line on each side al>ove lateral line; lower surface of 
head whitish, without crossbands; spinous and soft dorsal and caudal each with 4 dusky brown cross- 
bands; base of anal with 2 broad similar bands, and soft jwrtion of anal with 3 series of irregular 
crossbands; axil of pectoral above with white blotch; pectoral whitish with 10 blackish crossbands; a 
brown band in front of base of pectoral extending on lower pectoral rays; ventral with dusky blotch 
at base, outer portion with about 5 diL«ky crossbands. 

The only example we have seen of this species is the type, No. 50650, I^. S. N. M. (field No. 05030), 
6 inches long, obtained by us at Honolulu. 

48. Scorpsenopsis catocala Jordan & Evermann, new species. 

Head 2.1 in length; depth 2.75; eye 7.25 in head; snout 3.1; interorbital 4.3; maxillary 1.8; D. 
xii, 10; A. Ill, 5; P. 18; V. i, 5; S(-ales 9-42-22. 

Body elongate, greatest depth at first dorsal spines; Imck elevate<l, swollen, or convex, below first 
dorsal spines; snout rather long, with an elevated prominence; mouth large, oblique; maxillary large, 
expanded extremity broad, 6.5 in head; teeth in broad villiform bands in jaws, those on vomer small; 
no teeth on palatines; tongue small, pointed, free in front; lips rather thick, flt^hy; eye small, a little 
in front of middle of length of head; a deep pit below eye; top of hea<l with deep square pit just behind 
interorbital space; anterior nostril with broad fleshy flap; posterior large, without flap; four spines on 
side of snout above anterior nostril; preocular, supraocular, postocular, tympanic, parietal, and nuchal 
spines present; a series of spines running across cheek l)elow eye; several large spines on lower part of 
preopercle; several spines on opercle; side of head above with many small spines; suprascapular with 
several small spines; dorsal spines rather strong, third longest, 3.75 in head; last dorsal spine 3.8; 
second dorsal ray 2.7; second anal spine enlarged, a little longer than the third, 3.4 in head; first anal 
ray longest, 2.4 in head; caudal rounded, 2 in head; pectoral large, lower rays thick, fleshy, curved 
inward; sixth pectoral ray 1.7 in head, lowest 3.7; l)ase of i)ectoral broad, 2.25; ventral spine ^strong, 
3.1 in hea<l, second ray longest, 1.9; the innennost ray joined by a broad membrant* to l>elly; caudal 
peiluncle compressed, its depth 4 in head; head and body with many fringe<l fleshy flaps; scales mod- 
erately large, ctenoid. 

Color in life (fleld No. 03382), excessively mottled, streaked, and spotteil; body dark purplish 
brown or claret shaded, the spaces gray tinge<l with sulphury yellow; head all dull brown, flaps colored 
like the space about; belly to axillary region whitish with reticulations and irregular marks of yellowish 

^ =:5 



olive; axillary r^ion wine-brown, finely mottled with yellowish white in Rtn^aks and spots; a few 
round black spots l)ehind and in axil; inside of i)eotoral with a large jet-black blot<^.h at upjier part of 
base, bordered with orange; around this a largt* yellow area, then oblong black spots on the mem- 
branes of upi^er rays above middle, then a broa<l rose- red band, fading into violet below, the rim gray; 
ventrals bright brown and gray, red shadinl on inner face; inside of branchiostegals salmon-color, 
stripeil with white, the membranes yellow; memV>ranes (»f upper jaw salmon-color mottled with light 
yellow; tip of upper jaw orange with a golden ridge dividing a triangular spot of indigo-blue between 
vomer and premaxillary; a golden line on each side in front of palatines; tip of tongue light yellow; 
a triangular indigo-colore<l spot l)ehind teeth of tip of lower jaw; a golden streak l)ehind it on membnine 
before tongue; lower lip salmon-color especially l:)ehind where hidden. 

Color in alcohol, dark purplish, beautifully mottle<l with dusky and darker; head mottle<l alx>ve 
with dusky; fins with many fine dusky and brown wavy lines; base of pectoral both outside and 
inside brownish, the latter variegated with white and blackish brown; outer [>ortion of inside of 
pectoral covering first 5 rays with a series of broad blackish sjx)tf<; ventrals mort» or less brownish 
variegated with gray and whitish; Ixxiy whitish, mottled with i>ale brown; edges of butx»l folds, 
inside of mouth, deep yellow; a deep blue blotch directly l)ehind ti^eth in front of each jaw. 

This species is related to Scarpinia gihbosn (well figureil by Gunther in Fis<*he der Sudse(»)» 
from which it differs in the much rougher and letm depressed head, nnich larger flajw on oi)en'les and 
mandible, and the presence of a very large fringed flap on the anterior nostril, this l)eing» ol)solete in 
S. gibbosa, AVe have comixareil our sj^ecimeus with examples of *S'. gibbam from Apia. 

This species was obtained l)oth at Honolulu and Ililo, and appears to l>e not uncommon. Our 
collections contain 8 exc^ellent examples, as follows: 

















7.5 ; 




I • • • • ^« V ' • • • • 



Final dl.siMwiticm of .siH.H'imon. 

Type, No. 50651,U. S. N. M. 
Cotvpe. No. 74r>6. L.S..)r. Tniv. Mum. 
Cotypo, No. 2717, V. S. F.C. 
(•olype, No. 1698, Bishop Mus. 

C<>tvi»e. No. . M. v.. Z. 

Cotvpe, No. 3y6(i. Field Col. Mus. 
CotviH?. No. 2293, Am. Mu.h. Nat. IHst. 
<;otV|K', No. 21221, Ac. Nat. S«i., Phila. 

40. Dendrochiriifl hudsoni .Ionian & Kvermann, new sinnnes. 

Head 2.5 in length; depth 2.5; eye 3.4 in hea<l; snout 3.3; interorlntal 5; maxillary 2. 1 ; mandible 
1.8; D. XIII, 10; A. iii, 6; P. 18; V. i, 5; sirah^ 8-52-13. 

Body elongate, comj)res«ed, rather deep, the gn*atcst depth at fifth dorsal spine; profiles of trunk 
above an<l below more or less even; head com])re8sed; snout short, rounded; mouth large, maxillary 
nearly reaching l)elow middle of eye, it« distal expande<l extremity eijual to half eye; minute tiH'th in 
bands in jaws and on vomer; liiw thin; tongue jMjinted, ccnupressoil, free in front; jaws nearly CM^ual; 
anterior nostrils each with a small fleshy flap; interorbital space dimply concave; fifth dorsal spine 
longest, 1.25 in head; i)enultimate spine 5.2; set^ond anal spine longest, 2.1; thinl anal ray longt^st, 1.3; 
caudal rounde<l, 1.25; j^etrtoral 2.4 in trunk, ri'aching below middle of bam^ of soft dorsal, n)un<led, and 
only membranes between lower rays slightly incisiMl; ventral nninded, rea<'hing base of first anal ray; 
caudal i>eiluncle comi)reH8e<l, it« least <lepth 3.5 in heatl; nasal spines very small; preocular, postocular, 
tympanic and coronal spinels pri'sent; parietal and nuchal spines forming a single ridge; a ridge oi 
spines behind eye al)ove oi)ercle; a ridge of spines Im^Iow eye, ending in a spine on margin of pre- 
opercle; 2 spines l)elow this also on margin of pnM)iH»n*le; no opercular spines; margin of pn*oiH*rcle 
with spine projecting down and back; skinny flap alK>ve eye e<|ual to its diameter, and another from 
preorbital spine; scales small, ctenoid; hea«l nake<l except some seniles on oiH»rcle, cheek, and side 
above; lateral line running obli(|uely down to base of ciuulal. 

Color in aUx)hol, pale brown or whitish; side with 3 pairs of det^p brown vertical bands, first on 
posterior part of heaii prece<led by adwp bn)wn streak from 1k»1ow eye, stH*<md on middle and i)oeterior 
part of spinous dorsal, an«l thinl extending out on soft anal and basal iK>rtion of soft dorsal; soft 
dorsal (»udal, and anal pale or whitish; membranes of dorsal spines deeply incised in front, each 



spine with 3 brown crossbands; pectoral grayish with a blackish brown basal blotch and 5 blackish 
crossbands; ventral blackish with 2 whitish or grayish blotches. 

This species is especially characterized by the unspotte<l soft dorsal, anal, and caudal. From 
DendrochiniA Ixirf^eri Steindachner, it is distinguished by the longer pectoral which reaches to below 
the posterior dorsal rays. 

Named for Capt. C. B. Hudson, in recognition of the excellence of his i>aintings of Hawaiian 

We have examine<l 5 specimens of this si)ecies, as follows: 






' InchtM. I 
03W7 l.S Waikiki 

Final <lisiHwitir)ii of siK.K*iiiKMi. 

TyrM'. No. nOt'.r.J. r. j^. n. m. 

651 1.9 Korf near Honolulu ('otvj>e. No. 74r>7. L. S. Jr. I'niv. Mu.s. 

r»52 1. U Honolulu ('ol*vi)i*. No. 271H, V. S. F. C. 

O.P.J. 1 

aOl 1 do ' 

:J..5 do 

Family GOBI I D^. The Gobies. 
QUISQUILniS Jordan ft Evennann, new genni. 

Quisquilius Jordan & Evermaun, new genus of (iobikhv (engenhi^). 

Allied to Aster ropterifx. Bcxiy robust, covered with large, ctenoid scales; snout blunt; mouth large, 
very oblique, with 2 series of sharp teeth in jaws, the inner depressible; side of head with several 
series of short papillary fringes; ventrals sejxarate, their rays i, 5, joined at base by a narrow frenum; 
dorsals short, the first with 6 spines, the second with 12 short rays. 

The genus is distinguished from other small Eleotrids by the papillary' fringes on preorbital, jaws, 
and opercles. 

50. Quisquilius eug'eniiis Jordan & Evermann, new species. 

Head 2.8 in length; depth 3.8; eye 3.25 in heatl; snout 4.25; width of mouth 2.4; interorbital 2 
in eye; D. vi-12; A. 10; V. i, 5; scales 25,-12. 

Body robust, compressed, greatest- depth alx>ut middle of belly; hea<l large, elongate, broad, depth 
1.4 in its length, width 1.25; snout short, blunt, rounded alK)ve; jaws large, lower pn:>jecting; month 
large, very oblique, its posterior margin reaching Inflow front of eye; upper jaw with 2 series of 
teeth, sharp- pointed, outer larger, the inner depressible; mandible with teeth similar to those in 
upper jaw; no teeth on vomer and palatines; tongue tnmcate, front margin not notche<^l ; eye large, 
high, anterior; n<jstrils separated, anterior in small tul^, posterior close to upi)er front margin of eye; 
interorbital space narrow, very deeply furrowed; a series of fringe-like j)apilhe ninning from preorbital 
along upper margin of maxillary down Ix^hind comer of mouth where it joins another series nmning 
along under surface of mandible, and continue<l back and upwanl on margin of preopercle; anterior 
margin of open-le with a small vertical series of papilhe, eai^h papilla a little shorter than diameter of eye; 
gill-opening large, continued forwanl till nearly l)elow iH>sterior margin of eye; spinous dorsal rather 
small, spines flexible, with tips pnxluce*! in short filaments; soft dorsal high, median rays rather 
longer than others; anal more or le*« similar to soft dorsal, |)08terior rays very long; caudal rather 
large, round; pe<'toral broad, round, i»qual to hea<l; ventrals small, 1.25 in hciwl, sharp-pointed, and 
joined at base of inner rays by a narrow frenum; caudal i)e<lun(!le compressetl, its length l.(> in head, 
depth 2.4; scales large, ctenoid, those on upper part of head very small; snout, interorbital space, 
jaws, and K>wer surface of hea<l naked; no lateral line. 

Color in life (field No. 03554), Ixxly with transverse Ixands of dark brown with olivaceous tinge 
alternating with dirty white; edges of scales in dark brown j)ortions lighter; dorsal, anal, and caudal 
dark brown, edged in part with white; i)ectoral light reddish brown. 

Color in alcohol, brown; 12 dark brown crossbands on side, the last 6 very broad, much broader 
than the pale interspaces; vertical fins dark slaty; pectoral pale slaty; ventral pale on outer |x>sterior 
portion, blackish slaty on inner. 



We have examine<l the following examples: 






^ji^J^^ I length. I 


Disposition of spooimeiift. 

Waikiki .... 





Tvpc, No. f,0674. r. 8. N. M 
(\>tyiH;, No. 7483. L. S. .Ir Tniv. Miw. 
CotyiK'. No. 2732 IT. 8. F. C. 
(\»tviK', No. 170S, Bi^^ilOp MU". 

Cotype, No. . M Z 

CotviK.*, No. 3*r70. Field Col. Mu8. 
Cotype. No. 24230, Ac. Nat S<i. Phila. 

51. ^ Gnatholepis knig'liti Jonlan I't Kvennann, new Pi>e<'ieH. 

Head 8.5 in lenjrth; tlepth 4.2.>; eye 8.S in heivl; snout 8.<>; width of mouth 2.5; interorbital 
2.25 in eye; I), vi-12; A. 12; P. 1«; v! 5.5; scales 82,-9. 

Body elongate, compresfed, not «leprea*e<l in front, ^rc»atost dei>th at the middle of bell]?; head 
elongate, ita depth 1.25 in its length, its width 1.5; i^nout obli<|ue, blunt, brua<l; upi)er profile of the 
head obtuse, with a prominence over eye in front; mouth rather broa«l, the maxillary not n^ai;hing 
posteriorly to l)elow front rim of orbit; lips nither thin; teeth small, sharp, in narrow l)and8 in 
jaws with an outer enlarge<l sc*ries; no teeth on vomer or |>alatines; interor]>ital H]>ace very narrow, 
'evel; nostrils small, clo«» togt»ther in front of eye, anterior with Hap of very short, flt^shy cirri; 
eye high, small, a little anterior; gill-oiH*ning ri'stricted to side, nearly vertical, its length 2.25 in 
head; scrales large, finely (!tenoid, and becoming much larger on ix>sterior side of tmnk; scales 
small on In^lly in front of ventrals, cycloid; scales nuxlerately large, cycloid on the upper part and 
side of head, head otherwisi^ nake<l; dorsal fins well separate*!, spines flexible and with extremities 
of most fn.'e and filamentoiw; first 1.(5 in head, tifth 1.7, last 2.7; s<.»ft dorsal h>ng, la.««t rays longest, 
first 1.7, last 1.25; anal similar to the dorsiil, but lower, first ray 2.8, last 1.25; caudal n>uiule<l, the 
median rays very long, a little longer than head; ptn^toral with upjKT nu^lian rays longest, all 
rather fine, about equal to length of caudal; ventnils rather large, frenum uniting in front, rather 
Inroad, length e<jual to jKJCtoral; caudal i)eduncle compresstnl, length 1.2 in hea<l, depth 2.25. 

Color in life, j)ale flesh-color, ui)per i)arts with dark brownish spots and blotches; a st»rie8 of about 
8 brownish V)lot<*hes along middle of side; a small dark spot on base of pectoral; opercle dusky; fins 
all pale, spinous dorsal with brown edge; iris bluish white. 

Color in alcohol, pale brown, side with numerous small dark brown spots and 7 large dark brown 
blotches; a d'ark brown streak below eye, anri another across oj)er(4e; spinous dorsal very \\s\q brown 
with a'bout 3 l)lacki8h brown cross-lines, very distinct on first spines, running somewhat obliquely, 
and becoming indistinct posteriorly; soft dorsal with the spines i)ale or whitish brown and membranes 
between blackish brown; anal more or less <lark gray brown; caudal very pale brown or whitish, 
spotted in cross-series with brown; pectoral pale brown; ventrals dark brown, paler along e^lges. 

Color when fresh, of exam])le fn^m Hilo, olive-grt^en, rather pale, and with 7 bla<^kish crossliands; 
caudal sj)ot small and inconspicuous; black l>ar Ixilow eye, narrow and very distinct; Iwick crossl>arred 
with many spots of dusky olive; side with longibi<linal streaks of dark brown sj>ots along rows of 
scales, these irregular and variable, mixe<i, especially Ixrhind, with siK>ts of jxale sky-blue; dorsal, 
anal, and caudal dotted finely with dark olive; pectoral pale olive; ventrals blackish; anal plain 
bla(?kish, paler at base. In most examples examined the head was finely dotted with bright pale blue 
on cheeks and ojx^n'les. 

This small but interesting species is generally common in Ijrackieh water about Hilo and Honolulu. 
Our collections I'ontain a total of 128 si>ei'imens; 15 of these have been tagge<l and their measurements 
are given in the table; 101 other siK«<rimens from Hilo range in length from 1.1 to 2.5 inches, the 
average length Iwing 1.81 inches. From Waianae we have 5 specimens, 1.8 to 1.8 inches in length, the 
average Ixjing l.()2 inches. From the pond at the Moana Hotel at Waikiki, we have 2 examples, each 
0.8 of an inch long. The average length of our 128 spe<"imen8 w 1.81 inches. 

The si>ecies is named for Mtister Knight Starr Jordan, who first notice<l it in the i)ond at the Moana 
Hotel at Waikiki Beach near Honolulu. 

'» The jrenn-M OnatholepU Blcfki-r flecm.M to Iw equivalent to Ilazcu* of Jordan Jl Hnyder. 



The following are some of the specimens examined: 






2. l!n 









2. 2 












2 2 






















Final di.sjHudtion of specimen. 

Type. Xo. 50653. U. S. N. M. 

Colype, No. 7468, L. S. Jr. Univ. Mu». 

Cotype. No. 2719, IT. S. F. C. 

Cotvpi*. No. . M. C. Z. 

Cotv[>e. No. 2294. Am. Mu.s. Nat. Hist, 
Cotype, No. 24222. Ac. Nat. S<'i. Phila. 
C<:)type. No. 9812, Ind. Univ. Mus. 
CotypK?, No. .^1969, Field Col. Mus. 

Cotvpe, No. . 

Cotyi)e, No. 1497, Cal. Ac. Sol. 
Cotj-pe, No. 1699, Bii*hop Museum. 

Acentrogobius ophthalmotstnia. Streets, Bull. U. S. Nat. Mus., No. 7, 60, 1877 (coral reefs at Oahu); not of Blceker. 

52. Qobiopterus farciznen Jordan & Evennanu, new s^xndes. 

Head 3.25 in length; depth 3.5; eye 3.2 in head; snout 3.5; I), vi-11; A. 9; scales 28 (27 to 29)-10. 

Body rather robust, compre8se<l, greatest dejuh at gill-opening; head rather lai^, depth 1.25 in 
length, width 1.4; upper profile of hea<i evenly convex from tip of snout to origin of dorsal; jaws 
large, -mandible very large, slightly pro<luced; mouth large, very oblique, maxillary extending l)eyond 
front margin of eye; teeth in jaws unii5erial, rather large, somewhat canine-like; two small depressible 
canines on posterior part of bone behind anterior series; lips large, thick, fleshy; tongue not emargi- 
nate, large, thick, rounded; nostrils close together, posterior very large, in front of upper margin of 
orbit with elevated rim; interorbital space very narrow, concave; scales large, ctenoid; a large pore 
behind and above l>ase of pectoral; gill-opening large, continue<l forward Inflow; spinous dorsal small, 
flexible, spines ending in filamenti', beginning behind base of pectoral; soft dorsal high, rays of nearly 
uniform length; anal with posterior rays elongate, much longer than anterior; caudal elongate, 
rounde<l; pectoral broad, round, equal to head; ventrals long, equal to head, broa<l, without any 
frenum in front; caudal pe<luncle compret^sed, its length 1.5 in head, depth 2.25. 

Color in alcohol, pale brown, tnink covered all over with very pale minute brown dots; fins very pale 
brown, dorsals dusky, especially the spinous; 3 vertical pairs of pale brown cross-lines over side of head. 

Described from an exam]>le 1.1 inches long, taken at Hilo. Type, No. 50654, U. S. N. M. 

V IT It ART A Jordan ft Evermann, new genus. 

Mtraria Jordan <fe Evennann, new genus of (uthUdiv^ subfamily Tjuciogohiniir {clare^cens). 
Body elongate, tnuislucent, covere<l with very small thin scales; mouth small, oblique; teeth 
minute; gill-opening rather narrow; dorsals small, the niys v-11; jKH^toral rather long; ventrals small, 
unite<l in a circular disk. Small gobies of the coral reefs, allied to the Japanese genus Clariger, but 
with the first dorsal of 7 small spines instead of 3. 

63. Vitraria clarescens Jordan <& Evermann, new species. 

Head 4.6 in length; depth 6.7; eye 3.5 in head; snout 4.5; D. viii-11; A. i, 10. 

Body elongate, slender, compresseil, greatest depth between dorsal fins; head elongate; pointed, 
conic, depth 1.75 in its length, width 2; snout rather long, rounded; jaws prominent, upper slightly 
produceii; mouth oblique, maxillary reaching a little beyond anterior margin of eye; teeth not evident; 
tongue broad, truncate; snout alwve, interorbital space, and top of head more or less flattened; noetrils 
well separated, anterior nearly midway in length of snout, posterior close to front of eye; eye rather 
large*, anterior; gill-oj^enrng restricteil to side, rather small; scales very small; dorsal spines flexible, 
first dorsal small, the last three spines very small (minute stubs, broken in the type) the fin beginning 
behind tip of ventrals; soft dorsal beginning a little nearer base of caudal than tip of snout, about 



over insertion of anal, and anterior rays of l)oth finn longent, those of anal gnwhially smaller behind, 
the last 2 minute and close together; caudal emarj^inate, lol>€»8 rounded; ptH'toral rather lonjf, lower 
rays lonj^est; ventrals small, united to form a small round disk whose diameter is 2.25 in head; 
caudal petluncle compressed, elonjjate, its leni^rth e<|ual to head, its depth 2.25 in head. 

Color in alcohol, very pale translucent brown, 7 V-slia[>ed pale brown markings on up|)er side of 
body united over back; fins whitish. 

We have examined 7 examples obtainetl at Hilo, each about 1.2 inches in length: 




Final diHpoRitioii of Hpt'cimen. 

Typt\ No. 50jy>ft. U. S. N. M. 
Cotypo, No. 7469. L. 8. Jr. Univ. Mum. 
CotN-pc, No. 2720, IT. 8. F. C. 
Cotype, No. , M. C. Z. 


Final diHi)OHitlon of Hp<'cimcn. 

Hilo ! Cotyp*'. No. 2295. Am. Mm. Nat. Hist. 

I>o I 0»typ<'. No 24223, Ac. Nat. Sci. Phila. 

Do I Cotype, No. 1700, Bishop Muucuin. 




OBUBITB Jordan ft Evermann, new genns. 

Osurvs Jordan & Evennann, new genus of Pterojuiaridx (ParajnTcis aehaulnslamH Steindachner). 
This genus is allied to ParaperciSy from which it differs in having the caudal tin deeply forked 
instead of truncate. 


54. Fierasfer umbratilis Jordan & Kvennann, new species. 

Head 10.2 in length; depth 15.2; eye 5 in head; snout 4.8; mouth 2.6; interorbital 4.5. 

Body very elongatii, compressed; tail very long and tapering gradually in a long point; head 
elongate, c(mic, its depth 2 in length, width 2.25; snout rather broad, conic, and pnMluce<i l)eyond 
mandible; mandible broa<i, flattened lielow; mouth nearly horizontal, broa*l, the gaj^e reaching lielow 
posterior margin of eye; eye rather small, anterior, without eyeli<l, and i)lace<l al)Out tirst quarttT of 
head; nostrils well separated, anterior with elevated rim, posterior a short, crescent-like slit; inter- 
orbital space rather V>road, convex; gill-opening low, inferior, rather long; gill-membrane free from 
isthmus, its angle nearly an eye diameter distant from ix)sterior margin of -eye; dorsal fin aimost 
nidimentary, very low and thin; anal rather broad, in middle its height is al)out 0.75 in (»ye, from 
which point it grailually decreases to tip of tail, where it is rudimentary, Hke dorsal; tail ending in 
a fleshy point, caudal fin apjmrently absent; i>ectoral small but relatively large, 3.1 in head, rays very 
minute; lateral line distinct, running down along middle of side on i)08terior half of tail; no scales. 

Color when fresh (field No. OJ^'WK)), pale olivaceous, with pale greenish spots; a pale bluish streak 
in each spot over lateral line; pale purplish o])long s]>ots on lower half of Ixniy; head greenish-olife, 
with pale green spots closely set on chcH*k ancj jaw; pale puri)lish dots on upj)er part of cheek and 
behind eye; first dorsal same as bo<iy, but the sjHjts yellowish; a black sixjt l)ehind first and second 
rays, tips pale; rays of second dorsal checked alternately with yellowish-green and white; caudal 
same as second dorsal, but margin yellowish; anal, yellowish-olive; tip blackish; i)ectoral and ventrals 
pale; iris greenish -yellow; dull red stR^aks radiating from pupil. 

Color in alcx)hol, brown; head and end of tail dark sooty or blackish brown, the (!olor formed of 
dark points; greater part of anal fin, lower surface of lx)dy anteriorly and i)ect<)ral and branchiostegal 
membran*^, pale straw color; lower surface of trunk more or less blotcheii with pale brown. 

Our collection cxmtains but 2 specimens of this species, lx)th obtaine<l at Ililo. Tyi>e, No. 50656, 
U. 8. N. M. (field No. 03506), a specimen 7.6 inches long; cotype. No. 7470, L. S. Jr. Univ. Mus. (field 
No. 528), an example, 6.4 inches long. 

Fierasfer innftratUiit occurs also in the South Seas, and is readily distinguished from most related 
species by its dark, non -translucent coloration. 

R horahoreima from Borabora, briefly described by Kemp, has the pectoral 6 to 7 times m head. 


Family PLEURO\ECTID.€. The Flounders. 

55. Engyprosopon hawaiiensis Jonlan i^ Evennann, new sjic^'ic^. 

Head 3.8 in length; depth 1.75; eye 3.25 in head; snout 4.25; interorbital 6.3; maxillary 2.8; I). 79, 
A. 56; P. I. 10; V. i, 5: st-ak*?* 14-46-15. 

Body elongate, deep, rather ovoitl, greatest depth about end of pec^toral; head very deep, its length 
0.7 in depth; upper profile very convex in fn:>nt, steep; sn«»ut short, obtuse; jaws small, produfe<l a 
little, the mandible sliiihtly pn»jeoting; li[>s rather thin; mouth curved a little, very oblique, the small 
maxillar\' reaching a little l>eyond front margin of eye; tii'th in jaws very small, sharp- jiointeil; eyei« 
well separated, lower anterior, placnl in tirst thinl of head, the upper aV>out two-fifths an eye diameter 
posterior; nostrils cK^se together, with elevate^l rims; inten.)rbital spaiv a little moit* than half an eye 
diameter in width, deeply concave; gill-opening small; gillraker^ rather short; scales large, finely 
ctenoid, very small on rays of vertical fins; lateral line strongly an-hed at first for first two-ninths its 
length, then straight to base of caudal; don^l beginning on snout, the anterior rays free for only a 
short portion of their extremities, fin?t 5 in head, fiftieth 2.1, this the highest region of the fin; anal 
more or less similar, first 3.25, thirtieth 2; cau<lal rounded, middle rays longest, 1.1; pectoral short, 
pointed, 1.4; ventrals rather broad, l>aseof left 3, first and last rays aboat equal; right ventral smaller; 
caudal peduncle compresse*!, its depth 1.9. 

Color in alcohol, dark olivaceous brown, fins dark gray-brown, each ray finely specked with oliv- 
aceous brown; left pectoral specked with dark brown, right pei-toral dull creamy or brownish white 
like the right side of Ixxly. 

Type, No. 50657, U. S. N. M., taken at Ililo, the only example we have seen, 3 inches long. 

56. Bngyprosopon arenicola Jordan & Evermann, new species. 

Head 3.6 in length; depth 1.9; eye 4.3 in head; maxillarj* 3; D. 78; A. 57; P. i, 11; V. 5; scales 
14-36-17. Body elongate, verj' deep, rather ovoid, the greatest depth at tip of pectoral; head much 
deeper than long, the upper pn^file steep, strongly convex; snout obtuse; jaws very oblique, man<lible 
slightly projecting; maxillary very oblique, reaching below anterior margin of eye; lii>s rather thin, 
fleshy, fringed along margins; teeth in jaws minute, sharp-pointed; eyes close together, lower anterior 
placed about first thinl of length of head; upper eye alx)ut one-third an eye diameter posterior; 
nostrils well separated, with raised fleshy rims forming a flap: interorbital sjiace very narrow, concave; 
gill-opening rather small, restricte<l to side: gill rakers small, short, few: scales large, finely ctenoid; 
lateral line strongly arche^l for anterior fourth of its length, then straight to luise of <^udal; anterior 
dorsal rays free distally for one-half their length, first ray 3 in head, forty-fifth l.S. which is the 
highest region of the fin; anal similar to dorsiil, but anterior rays not free for half their length; first 
ray 3.5, thirtieth 1.8; caudal elongate, median rays longest, e<]ual to head; pectoral short, pointed, 1.5; 
ventrals rather large, the left with its base 5 in head, first ray 3.6, last 2.6, almost entirely in front of 
the right, which is much«emaller; i-aujial pe<iuncle broad, comprt»<se<l, its depth 2.2 in head. 

Color in alcohol, very pale brown; side marke<i with many large incomj>lete rings of blackish or 
dusky and with a number of dusky spots in between; fins whitish, the vertical or un])aire<l with large 
blackish spots on membranes between rays and similar small ones scattered! al)out, those of caudal 
forming about 4 crossbands; several dusky spots at base of pectoral; right side whitish. 

We have seen but 2 examples, l>oth taken at Hilo: Tyi>e, No. 50658, U.S. N. M., 2.5 inches long. 
Cotype, No. 7471, L. S. Jr. Univ. Mus., 1.9 inches long. 

Family .\XTEX\ARIID.€. 

57. Antennariois drombus Jordan <fe Evermann, new species. 

Head (to end of opercle) 2.5 in length; depth 1.75; eye 5 in head; snout 4; width of mouth 2; 
D. I-I-12; A. 7; P. 12; V. 5. 

Bo<iy very deep, compressed, l)ack elevated; head deep, with blunt conic profile in front, Si»me- 
what oblique above; snout broad, obtuse, surface uneven; mouth brx»ad, large, nearly vertical; maxil- 
lary concealed under skin, reaching below anterior jiart of eye; lips fleshy; teeth in jaws minute, in 
narrow bands; teeth on palatines rather large, sharp-pointe^i, none on vomer; tongue broad, thick; 


mandible lai^ge, with fleehy knob at symphj'sifl, projecting; nostrilB circular, well separated, with 
rounded fleehy riins; interorbital space tx)nvex, roughened; top of head with rather large concave pit; 
eye high, anterior; bait rather short, only reaching a little beyond first spine, with fleshy caruncle at 
extremity; dorsal spines short, first free, rough, depressible in pit on top of head; second doraal «pine 
twice length of first, etiual to width of mouth, depressible, and united with skin of back to its tip; 
posterior dorsal rays longest, and the last, like that of anal, united to caudal peduncle by a membrane; 
anal similar, rounded, elongate, }.5 in head; [K'ctoral broad; ventral small, rounded; caudal i)eduncle 
small, compressed, its depth equal to interorbital space; body rather rough, mucous pores on head and 
in lateral line with excrescences; side of Ixxly with many pointed cutaneous flaiw; second dorsal spine 
and first dorsal ray very rough, also with cutaneous flaps; lateral line very convex, running down 
toward middle of base of anal. 

Color in alcohol, pale plumbeous gray, more or less spotted or mottled with darker; Ixjlly and lower 
surface rather pale, the spots distinct; fins all more or less pale with dark spots, some at basal por- 
tions of dorsal and anal darker; iris blackish with radiating lines of golden. 

The above description is from the type. No. 50659, U. 8. N. M. (field No. 541), taken at Waikiki, 
near Honolulu. 

Another example (field No. 539) shows some differences: Head (to end of opercle) 2.5 in length; 
depth 1.7; eye 3 in head; maxillary 1.8; width of mouth 1.7; interorbital 3.7; D. i-i-12; A. 7; P. 12; V. 5. 

Body very deep, compre«se<l, Imck olevate<l; head deep, gibbous, with blunt conic profile in front, 
somewhat oblique al)ove; snout broad, obtuse, short, surface uneven; mouth large, obliquely vertical; 
maxillary large, reat^hing a little beyond front portion of eye; lips fleshy; teeth in jaws minute, sharp, 
in bands; teeth on roof of mouth large, sharp-pointe<l; tongue large, broad, thick; mandible laige? 
with knob at symphysis, projwting; nostrils well separated, close to end of snout, each with elevated 
fleshy rims, the anterior higher; interorbital space broad, elevated, uneven; top of head with rather 
large pit; eye high, anterior; bait short, reaching tip of first dorsal spine, with caruncle at extremity; 
dorsal spines short, depressible; first dorsal spine half length of second, free, depressible in pit on top 
of head; st^cond dorsal spine large, joined by skin to its tip; dorsal rays of about equal height, seventh 
1.3 in head, and the last, like lower portion of last anal ray, a<lnate to caudal peiluncle by a membrane; 
anal roundtMl; caudal elongatti, roundeil; pectoral broad; ventral small; body rather rough, mucous 
pores on head and lateral line with excrescences; along the lateral line and anterior region of dorsal 
are many <Mitaneous flaps; lateral line convex, running down to above middle of anal. 

Color in alcrohol, dark gray-brown; edges of vertical flns whitish, the pale border rather broad 
and very distinct along posterior, dorsal, anal, and caudal rays; side with alx)ut 6 lai^ round 
blackish siK)ts; caudal with some j)ale or indistinct mottlings; pectoral and ventral with rather broad 
margins, median i)ortion dusky; iris more or less silverj'. 

A. droinfnis seems nearest related to A. nummifer Cuvier & Valenciennes, originally described from 
Malabar. Prolmbly the specimens from the South Seas referred to the latter belong rather to A, 
dronihm, A. nummifer is said to be red in <rolor with dark spots, and, as figured by Dr. Day, differs in 
several respects from A. dromhiis. Both these spcM'ies differ from A, commernoni and its numerous 
allies or variants (A, niger A, lejyrom*j A. nihroftutnjin^ and A. mndrirensis from Hawaii) in the shortness 
of the first dorsal spine or fishing rrxi. This is scarcely longer than the second spine in A, drombus, 
but in A. commersoni it is twice as long. 

Our collections contain but 2 examples of this 8pe<ries, the type, No. 50659, U. S. N. M. (field 
No. 541), and cotype. No. 7472, L. S. Jr. Univ. Mus. (field No. 6Sd), both taken on the reef at 
Waikiki, near Honolulu. 




Siiirr till' publication of our recent paper" on new sjx'cies of li>lies from the 
Hawaiian Islands, further studies of our lart^e collections have resulted in the dis- 
coverv of an interesting new species of IrnpiiJn'lttlnjs and a rcmarkahle new ^enus 
of SmrjMi'n'nla . These arc described in the pn»sent pai)er. Illustrations of both 
8i>ecies will be jjfiven in our final rejH)rt. 

Tropidichthys psegma .It>r«Ian iV KvLTiiiann, iww >;|H*rit's. 

Ilfa«l :> ill leii^'th; di-pth 1, vyv 4.o in hviu\\ Miuut l.">; intrroHutal 'l.:\\ D. 11; A. 11; ('. S; 1*. U>. 

IUmIv nhort, stout, !inMl*'rately <"oni|>n*s.<4Nl: sriont Ion;;, conic-; anterior don^il profile ri>in>r<*venly to 
rcjrion alx»vi* jrill-oiH-nin^, at which iH>int iIk* ImmIv is <lt*<'|K*st; interorhital Hat; jrillH>jH*niiig nearly wr- 
tifal, short, its length less than iliaineter of eyt*; month low, U*low axis of IkmIv; tivth stronj;, convex, 
cutting' e<l^' sharp; eye small, supraorbital rim not prominent; i-andal |KNhinele <lei*p, it** lea»<t depth 
atM)nt 1 in head, its leiu^t width 4 in its least dej»th; len^rth of caudal jKMlunele from dorsal tin to l»aA* 
of eaudal lin l..*i in head: from Ut<<' of anal tin '1 in head; dorsal prominence e<|ually distant iH-twi-en 
tipof snout and |Histerior hasi*of <-audaI; Inis** of dorsiil 1.5 in height of lin, which latt«'r is 1 in head; anal 
similar t«» dorsal, it^ etlye roundnl; <au«lal truncate, or very sli^rhtly <onvex, 1.2 in head; |KH'toral 
hn>ad, its lia.*^* L'.*> in hea<l, fnn* etl-jc oMi.|ue, iM>sti'rior rays 1.') in anterior ont»s; ImmIv mostly smooth; 
int<'rorhital space and snout ahove and on side with small prickles; In'lly with a few prickles; a si-at- 
terinl patch al.*^» on >id<*alH»ve jK*ctoraI. 

Color in alcoliol. dark brown alHivc, |»aler lu'low; ;i or 4 short hla<k lin<-s runnim: forwanl ln»m 
orhit, and siinu* nundnT hackwanl; lower part of side. es|KM-ially p<.steriorly, and lower part of 
caudal iKihm<le. with small roimdish Mack «|M.t.<; snout and inten>rhital spa<*e <*rossed hy alMiut 12 
narrow hlack lines, thex* extending' down on sid*- of ^nout; >ide of snout with '> or 4 narrow l^lack 
lines from chin toward eye, separate<l ]»y j»ah"r lines; |M»sterior to thes** >mall irn^rular hlack s|M»ts 
cr>verin^ entire cheek, d<»tted ov«*r with line white sinnks; end.- of sj»ines, ]MM-toral, dorsal, and anal 
j>ale %\hitisli. their Kuh-s lar^rely hrownish hlack; caudal dark hrownish or hlack. 

This s|>ecies is known to us only from the ty|M', No. ."wiss"), I'. S. N. M. i Held No. 2.'Wil ) ;;.7.'» incht's 
lon^. ohtaiiunl hy us at Honolulu in IIHM. 

\Vi' have couijuiumI this si>e<*imen with examples of T. rnnmntn.^ Vaillant iV Sauva^re, ohtainiMl hv 
tlie AU>atn>si< in 11H)2, and tind them <|uite distinct. 

I&ACTJNDUS Jordan ft Evermann. new genus. 

/r/i/"/<r/»/.s .Ionian iV Kvermann. new urenus of .si-o//*.*/**'/.! ^sujinj,r\. 

Allied to lltlirttlntiis i\\\\\ I'mitums. ImmIv rather elonir:ite, compn*s.«HNl, ('oven"*! with small, wejiklv 
ctenoid nales; tins not S4aly; head not depres.^i'd; formed as in .'v //*/.</' ;<// x, the spines moderately 
develo|KMl; head and ImmIv with dermal flaps; teeth on jaw> and vomer, none on palatines; dorsiil (in 
deeply <livide<l, the spim^ 11 in mnnU'r, tlu* fourth nnich elongate; jK-ctoral rays undivideil; anal 
rays in, '>; ventral niys i, ">: caudal rounde<l; vent at Ikum* of first anal spine; air hladder ol»Hilet4\ 

'I UcMTiptioii"* of iirw iri'iicni ami vi»crn> of li>JK'> from the liuwniian I*«lainl>i. <Bull. I'. S. Fi>«h Oiuiin. VATl (April II. 

F. c. B. liW2— 14 20i) 


IracunduB signifer, Ionian & Kvi'rniuiiii, new H}K.*cie»<. 

lIoa«l '2A in U'n^tli; depth ',\.2; oye 4 in liea<l; niuxillary L': I), x. i, 1»; A. in, 5: 1*. 17; V. i, 5; 
wah'S alxjiit 11-55-:^), alx)nt An \hh\^. 

B<Mly rathor c*l«»n;!aU*. nuHirrattily foniiin-ssi^l, tho l»t'a<l i-oni**, nut «U»pix's>H'«l; month laiyi*, 
o]»Hi|ue, th** IdwtT jaw slijihtly projin-tin^, the niaxiliary reaching' to opjMisitr iM»sti*rior niaivin of 
pnpil; t^H'th in nKHh'rate ban<ls in the jaws, tlie inner tifth in the npinT jaw slij^htly hir^rcst; vomer 
witli Hmall te«.*th; jialatines tooth lesn; interorbital an'a <h*eply eon<-ave, little wi«ler than ]»upil; fpiiie.*< 
on t^»p of heail low and rather sharp, nun-h as in S^•h^l.stl^ffrs; pn^KMilar, snpiiuKMilar, iN>stoenlur, tyni- 
|>iuii<', <M*eipital, and nutrhal spines pres<.'nt: a ridjre with 2 spines ontside the tymi»tinie spine; [ireor- 
hital nuxlenite, aUuit as wide as eye, with a sharp s]>ine tnrned forwanl an<I a ])lnnt spine tnrned 
Inu'kward; MilH)rhital stay a narrow, simple rid^'e, ri'achinj: Imsj' of preo|K.'nMiIar spine, whieh is 
strai^rht and vt?ry short; .'> lower pn*oj>erenlar sj)ines n-dntM^l to hlnnt i>oints; ojKTeh* with 2 slender 
diver«rin^ spines, tlie np|K*r tlie lar^'r. their |M)ints not n'aehinjred;re of m<>mhran(^ head with nnmer- 
i»nH, ])roa<I, Heshy ilai)s; a frinj^d Hap at the nostril. 2 on i^\^v of pieorhital. It on lower limh of 
pr(^o]M'n'le, an<l a hi«rh frin^-il Hap a)>ove eye, alN>nt as lon^ tu' pnpil; small >imple ila]is on the ehi'ek. 
the end of tlie maxillary, an<l elnewhert* on head; lar^re pon's on lower jaw. nnder snlM>rhital stay, and 
elsewhere; operele and upiKT jmrt of eheek witli rn<limentary, emiNfldei] sj-ah's; jaws naktMl; top nf 
lii-ad sealeless, «Kripnt covered with thin skin ami seanrly depn-ss^Ml; pi 1 rakers very short, thiekish, 
and fee])le, all hnt aUait six n*<hieiNl to nuTe rndiments; no slit Indiind last ^Hll; ImmU' eoventl with 
small, elose-si't scales, whirli are sli^rhtly ctenoid; scales on nape small, on hri'ast minnte; lateral lim- 
couHpicnons, provitled with dermal tlajM; nnmcpMis si-atten**! flaps on sitle> of luwly. 

Dortfial Hn very <U"eply notche<l. the spina's rather slendrr, pnn;rent, tin* Mr>t a little lonp'r than 
eye, the wtrond ami thinl xilH'ipial, alnnit half lunger, the fourth jxn'atly eh'vatitl, l.Ti in hea<l, almost 
twiit^ heijfht of third ami fifth, which are siiInhiikiI; sixth, S4*venth, anil ei<;hth ^li<;htly lon;r<'r than 
fifth, tenth very short, eleventh lialf len;rth of fourth; soft dorsid hiyrh, the h>ni;est rays nearly tialf 
heiid; rays of all the fins scalek'ss; candal lon^% rounfleil. 1.4 in head; anal hi<;h, the spines ^'niduatitt, 
the thinl a little longer than siH'ond, which is 2.i> in head; loni;est soft mys l.,s in head; {Nftoral with 
the mys all simple, tin* hm^est 1.- in heail, lowi'st rays shortene<l and thicken^il; ventral tins inserteil 
U'low axis of jKH'toral, rather Ion;:, l.(» in liead, not ipiite rrachin^ anal, innj-r rays well fn-e. 

Color, |iale in alcMihol, donhtless vermilion re<l in life, the flaps on ImnIv pinkish; asin;;le jet-hhirk 
sjMit alN>ut half diamet<>r of pnpil near tij> of memhranr l)etw4'i'n s(M*ond and thinl spinrs of dorsal. 

The only example known i- the ty|K*, No. TK^NSii, \'. S. N. M. (field No. U'l")), a siM'cimen 4.2 inche.** 
lon^, taki?n hy ns on the coral reef at ilonolnln. 


• I 

TiiH frhsii-\vati:k iisiii:s of whsti:rn ciba. 

My C If. KICiKXM A.1SN, 

/*9tt/rs.\nr of XiHthtj^y, University of Imiiutm. 









Professor of Zoology, I 'niversity of Indiana. 

During March, 1902, the writer, aceompaniod b}' one of hi.s students, Mr. Oscar 
Riddle, as assistant and interpreter, made a series of collections in the fresh waters of 
western Culm, in the streams accessible by the Western liailway and the Unit^ 
IIal>ana Kailways. Attempts to reach waters remote from the niilwavs were alian- 
doned on account of the expense, })oth in time and money. Sumidero was reached 
b}^ horse from Pinar del Rio, and the caves about Caiiason foot and by volante. The 
orijifinal and chief object of the visit to Cuba was to secure material for a stud}' of the 
eyes of the blind tishes, Sfyfjimhi and Lucifurja. In this 1 was successful. The 
fresh-water tishes proved also of considenible interest. As mi<^ht have l>een expected, 
many of the more abundant and larji^er species had l>een previously descril>ed by 
Poe}. Nothing, however, was known al)out the distriV)ution of fresh-water tishes, 
and there were found a surprising number of new sj)ecies. 

I wish here to express my thanks to Mr. A. P. Livesey and Mr. J. E. Wolfe, 
the managers of the Western and United Ilabana Riiilways, who did all in their power 
to make the available time profitable from a scientific standpoint. I am also under 
obligations to Mr. Philip Hammond, the chief engineer of the United Ilabana Railroads, 
for suggestions and various favors. Mr. Pascual Ferreiro, of the CuIkiu railway 
jK)stiil d(»partment. kindly acted as guide to the Pedregsiles caves, and Mr. Francisco 
Martinez and his brother to the caves al)out Canas. The success of the expedition 
was largely due to my companion, Mr. Oscar Riddle, whose previous stiy in Porto 
Rico and trij) to Trinidad and the Orinoco had familiarized him with the language of 
the couiitrv and enabled him to d<»al with the natives. 

The dniwings illusti*jiting this paper were* made }»y Mr. Clarence Kennedy. 

In his •"Memoriixs sobre la Ilistoria Natural de la Isia de Cuba," tomo 2, pp. 
95-114, ISii^;, Poey describes two species of blind tishes. Lurifuf/n suhf cm incus 2ir\i{ 
Lui'lfHtjn (h-ntatuM^ from caves on the soutli<M*n slopes of the jurisdiction of San 
Antonio, Guanajay, and San Cristobal." They were tirst brought to notice by the 
sun'eyor, I). Tranquilino Sandalio de Noda. Specimens were secured for Poey b}' 
Dubroca, Fabre, and Lavunta. 

♦iJonlan i\: flvermnnn, in their Fishes of North niul Middh* Ameriea. iii. p. 2.t(I1, pive ft number of l<M-ulitie8 for 
which I can find no authority in Poey. ThiLs**Siin Antonio. Culia (Coll. I>. Tninquilinoi: Sandalio de No<la (('oil. D. 
Juan Antonio Fabrej"' and caveat the '* Castle of Concorrl." I can not find tlu' authority for the hx'ality San Antonio, 
Cuba; the collector given as D. Tranquilino is probably 1). Tranquilino Sandalio de No<la; the second UK'ality Sandalio 
de Noda is proUibly the latter s<'<*tion of the name of the man who first called attention to them. Castle of Concord 
should probably read " Coffee plantation I^ Concordia." 



The localitios from which Pory sociuhmI Lmtifinjn tiuhfrrrtin* ils jin» (1) Cave of 
Cajio, 5.2 miles south of La (jriiira (h» Moh»na (Noda, IS.Sl); (2) cave of the cofff»o 
plunttition La Indiistria, halfway hotwi^on Al<[uizar and (nianimar (I)u!)roca); (8) 
cavo of Ashton, noar San Andres (P^hro); (4) rave of th<» I)ni<^on, on the cattle 
farm San Isidro, near Las Man<pis (Fahre); (5) cave on the cotrcH* plantation I^ 
(J<)n(»ordia, 5.2 miles from Alquizar (Layunt4i); («J) rave near th<» heehouse of the 
cotfee plantation La Paz (Diibrcx^a); a well near the tav(Mn Frias (Fal)re). 

Luclfu(jt( fl//tf<fftf.s was secured in Nos. 1, 2, and 8. Tho-^*' from Nos. 1 and 2 
were said to l)e without the least vestiges of eyes; those from No :5 with vestij^es. 

1 visited a number of the cav(»s from which Poey secured his sjM^cimens, intendinjij' 
to ol)tain spe<'imens from as many of Po(»v's localities as possible, but especialh' from 
those from which he secured both species of the l)lind lishes. Th(^ towns (luira de 
Melena, Ahpiizar, San Andres (now ( -anas), mentioned }»y P<m\v, are suc<*essive sta- 
tions alon^ the Western Itiiilroad, and Las Man<ifas is a town a short distanc<» off the 
railroad Ijeyond Oanas. We made our tirst stop at Alquizar, hopinjif to l>c in the 
middle of the cave region. The drivcM- w(^ enjifji^ed at Ahpiizar to take us to Iai 
Industria showed his independ(Mic(» and orij^inality by takint^ us over a rouj^h road 
durinfif a half day's drive parallel to th(» railroad to Frias and Ashton near the station 
Cailas, but wo were not inclined to cpiarrel with him, as we at once secured tishes in 
Ashton, and caves were rej>orted to us as very abundant in the* whole refrjon south of 
Cafias. Many of these caves were visited hy us, as well as that at Cajio (l»y Mr. 
Riddle) and others in a widely distant j)art of th(» island. Those from which we 
secured specimens I propose bri(»fly to describe* here. Th<» cave on Lii Industria wo 
did not visit, and in fact, except while we w<»re en^a^ini^ our volante, no one about 
Alquizar seemed to know the plantation La Industria. Jt is |)ossible that the name 
has l)een chan<^ed in re<»ent y(»ars. Bearing in mind our experiences with the volante- 
men at Ahpiizar we mad<» no attempt to tind I^a Concordia, which is also reiu*hed 
from Alquizar. 

The ^'caves'' about ('anjis can !K\st be described after a few words concerning the 
country' in general in which they occur. The territory alM)ut (-anas is entii'ely 
drained by underj^round streams. The streams risinjj^ in the hills and mountains 
forming the watershed })etween north and south drainage run above j^round for a 
distjince and then disapjn^ar underj^round. The AritjuanalM) River thus runs into a 
])ank at San Antonio di* los Banos and disjippairs amon<)f fallcMi ro<*ks. A few vards 
away from its '\sumidero"' the water can 1m» seen runnintr in its underj^round channel 
thi'ou^h an ojH'uinj^ in tin* thin roof of the chann<d. A few yards farth(»r on a drv 
channel leads down to the water which at the end of tin* chann<d disiippeaiN amon^ 
fallen rocks. Other rivers disa])pear in a similar mann(M*. Th(M'r watiMs reap|)ear, 
in part at lejist, in a numb(»r of "'ojos,-' some near the coast south of San Antonio. 
The region drained by und<»r^round streams is flat, with frcijuently no indications of 
surfa<*e streams and their erosion, and extends westward to n(»ar San Cristolial^ 
where the first permanent surfa^'c stream is ()l)serv(»d. At Artemisa and Candelaria 
stream beds cont:iined pools of water at tin* tiim* of our visit. 

From San Cristo})al to Pinar del Rio there were many small but {MMTunial 
streams. Eastward from Cailias the cave ret^ion has an unknown ext<»nt. Poe}' 
limited it to the jurisdiction of Guanajay, but it certainly' extends as far cast as tlic 


m(?ridian of Miitanzas, and from reports probal>ly beyond Cienfuegos. East of Rincon 
then* are, however, fro(|uent river l)eds, dry during the time of our visit. This 
main rej^ion, lielonju^inji^ to the southern slope, sends a iX)int northward from Kineon to 
the Ahiiendares River in the northern watershed. Aside from the ''Ojos de agua" 
alonj; the ed<(c of the* ri(Mie<(as skirtin<>'the southern eojLst there are two notaWe places 
where underjifound rivers find an exit. One at Vento supplies the entire city of 
Hahana with its water, the other serves to make th<^ re<^ion al)out (fuines a garden, 
its waters Iwing used for irrigation. C)thers in all prohahilit}^ have a subacjueous 
exit to the south. 

The large spring at Vento, sufficient to supply the city of Ilabana with its water, 
is the only one which issues on the northern slope, so far as I know. The origin 
of the suppl}' from the Vento spring has not been traced. It issues but a few feet 
from the Almendares Riv(M\ and it is very remotely jx)ssible, though not at all prob- 
able, that it derives its water from the upper courses of the Almendares. At the 
time of our visit the water of the spring was 1 degree warmer than that of the 
Almendares River. The region north of that river, across the river from Vento, 
l>eing shut out from a possible contributing source, it undoubtedlv derives its water 
from the system of underground streams mentioned above. An examination of the 
lM\st available map and the levels of the Western and United Ilabana Railroails makes 
it seem quite certjiin that the Vento springs deprive their water from the region imme- 
diatelv south of Vento and north of Rincon and Iknucal — that traversed bv the two 
niilroa<ls mentioned. This region contains various sinks without surface outlets, as 
well as dry sink-holes, and is the northward-projecting {loint of th(» cave region men- 
tioned alxne. A notible sink-hol(» in this region is that at Aquada, on the United 
Habana Railroad. This is very broad, shallow, and dry during the*, drv season, but 
the water rises to stind over 10 feet deej) on the railroad tnick during some of the 
wet seasons. 


The soil over the region undiM- considenitiori is thin, the surfaces of the very irreg- 
ularl}' corrodefl rocks jutting out in lumierous })laces. This, together with the fact 
that the wat<»r of the unch^rground streams is but a few feet underground, givers the 
region an entin»ly dilferent :ispect from the sink-hole and underground-stream region 
in Indiana, Kentucky, etc*. It is, in the first j)lace, impossible to enter the under- 
ground streams, and there are no funnels on the surface to indicate the location of 
an underground stream or its tributary. In i)laces the thin limestone roof of an 
underground chamber has given way and enables one to get to the water, which in 
all the numerous places we visited was stationary, not flowing. With one or two 
exceptions the water was covered with a continuous crust of carbonate of lime, due to 
the evaponition and discharge of carl)on dioxide from the surface of the perfectly 
quiet lime wat(M*. When the water is distui l>ed flakes of variable size break loose 
and irniduallv sink to the bottom. All of the so-called caves al)Out Canas, WMth one 
possible excej)tion, were sink-holes formed by the breaking of the thin roof of a 
larg(»r or smaller underground chamber. In all of the caves where sbilagmites and 
stalactites were noticed these extended for 3 feet or more into the watcu*. Inasmuch 
as they could not have fornwd under water, the latter must have risen since their 


fortnation. Usually die walls of the sink-hole retreat downward and, 
su^ifestinj^ that a channel tilled with water extends down and out from the sink-liole. 
The impression is irresistibly made that there exists here an extensive series of 
drowned eaves. As our visit to the eaves oecurrc^d near tlie (»nd of th(» dry season, 
this exeess of wat^^r could not have becMi <hie to recent excessive niins. The water in 
the ciives was, however, still fallin*^ and fell sevenil inches during March. 

The condition could have been brouj^ht about (1) by the risinji^ of the southern 
coast of C/uba, resulting in a chan*^ed incline in the underground streams and a c<m- 
se^juent drownintjf of the ca^-es; ("2) by a blockinj,^ of the outlets of the streams; (Ji) 
by an increase of the rainfall above that i)rovided for in the past. I wjls at tirst 
inclined to favor the tirst view, since caves as widely diff<»rent jts those* of Canas and 
Ahwrnines showed the same diameters; but a cave at Matunzas, on the northern sIojm*, 
showi»d exaetl\- the same conditions. A local block injif of the outlets is scarcely 
conceivable on account of the wide separation of caves sliowinjf the sjime conditions. 
Concerning the rainfall 1 am not able to speak, but any other cause, since the caves 
are above the level of the ocean, docss not occur to me. 

The detailed description of various caves may bejfin with Mod(»sta No. 1. The 
caves, for tin* most part, do not hav(Mlistinctive names, l)ut are called aftc»r the Fin<*a 
on which thev occur. I have added the numlNM's 1, 2, (»tc., for convenience. Modesta 
No. 1 is an i(l<»al cave of which all the rest are moditications. There is nothint^' on 
the surface* to distin^ruish it when one is a few fe(»t away. The cave is b(»ll-sliaped, 
with an openin*'' 10 to 15 feet at the top. A tree throwing at its margin sends vertical 
roots down to th<^ bottom. On these roots notclies have been cut, and the descent is 
made by means of them. At the water level at the time of our visit the cave was 
oval in .section, IM> bv 45 feet in extent. In the middle of th<* bell, and immcdiattdv 
under the oixMiing, then* was a larjre pile* of rocks, cemented totfethcM* in places by 
st4ila<^mitic material and rising but a few inches above* the* water. The water, beauti- 
fully cl(*ar, became raj)idly d(»e|HM' in all dinn-tions and could be seen to extend out 
in at least two directions in (leejx^ninj; channels tilled to their tx)p with water. The 
roots desc(Miding from the ()p(*nini«f at the top to the island, a distance of about 15 
feet, h(*re dividenl suddenlv into a tuft of innumerable rootlets, most of them in the 
water. Such roots w(M-e found in almost all the cuves, and the younjif blind lishes 
were always found in amonj; the rootlets; the bio^ on(*s among the rocks. 

M(Klesta No. ^, also calhnl Hawey, is a cave of the same "type*, except that the* 
central mass of fallen re)e*k fe)rms an are'li over the* water, anel that it can Im> re>ae*lu*el 
from erne side* by an ine'line*d plane*, also forme>d e»f falle*n mate*rial. Keie»tlets we*re 
very abundant here* and small blind Hshe»s e*(jually so. The wate*r was pre)bably ne)t 
more than 10 fe^.et be*le)W the* surfae-e and at e»ne e*elge* was ve'ry ele'e*p— he)w ele*e*j) we 
had ne) me*nns e)f de^te'rmining. Part of the* e)[)e'ning hael at e)ne* time been walleel in 
and the e*ave was use*d as a well. It is j)e)ssible* that this is the* well me*ntie>ne*el by 
l\)e*y Jts Ixiing near Frias and contnining blind tishe*s. Hut a nuinlR*r of either wells 
in the ne'i^^hbe)rhoe)d re»ach e*ave»s. 

Modesta No. 8 ce)ntnined no blind tishe*s. 

San Isidro No. 2 is a duj)lie*atie)n of Mode^sta Ne). 1, ])ut with ditferent pre)pe)r- 
tions. The centnil mass is higher and heilds a numbe*r of stala<(mite*s. It is emly 
partially surrounded by wat^M*. The e*ntrance is })v re)ots. 

Bull. U. S. f. C. 



San Isklro No. 1 is an undorii^roiuul (rhanihi'i* containing 2 to 3 feet of water in 
plat-OS and nuiddy in tlio liottonh It is roachod by an inclined plarn; from the side 
and is more like a cave as ordinarily understocKl than the cjives of the Modesta tyjK\ 

La Frias No. 1 is a cave of the Modesta tyj)e, hut larger and with the r<K)f of 
one side fallen, so that th<* descent to the water is made by an inclined plane, and the 
water forms a crescent about l.^ feet wide under the retreatinjr walls of the side 
opposite to the entrance. 

La Frias No. 2 is another cav<' of the Modesta ty]H', loo f(»ct across at the lK)t- 
tom and with tin* water 40 feet from tlu' surfaces Th<* roof has fallen in at one side 
and the central mass of rocks rises nearly to the level of the surroun<lin^^ region, so 
that the descent can l)e made by a windin<5 inclined plane. Part of the roof, very 
thin and worn throujifh in places, is <till standinjr. suj)|X)rt<Hl by stala<^mite-stalactite 
<*olumns. The roots of trees wind alM)ut stalatrmites or descend as stniiji^ht and 
unsupi)ortx^d stems 15 and more feet throujrh the cave to the water, where there is 
the usual breaking up into rootlets. 

Ashton No. 1 is a cave of the Modesta tyjM». with a lart«fe amount of fallen mate- 
rial at one side, where oiKM'an descend to the wat<T by means of st<»j)s. The diameter 
of the cave is p<>ssibly HH^ feet. A lar<re tree tji'ows in the center of the fallen 
mat<>rial. At the deej^er part of the cave, oj)|)osite th<* entnince, the walls retreat 
downward. Stalactites and stalajnuites are j)rescnt, and there is a cn»sc(»nt of pools 
of water of variable* width and depth. Channels tilled with wattM* are soqii to lead off 
from the }>ottom of the ]X)ols. This is one of the rare jdaces where ^reen plants are 
fjrowint^f in the water. They an* contined to the dextral ik>o1s as one enters. 

Ashton No. 2 differs from all the other caves. It is more oj)en: the roof has 
fallen in, so that then* an» no rctreatin*5 wall>, as in the other caves. The fallen 
material sIo|K's j^nidually to the wat(M*, which is shallow and densely covennl with 
duckweed. A channel leads off from the left <^nd of th(» wat<»r as one enters, and 
we went into this with our canoe, but th<' waIN soon came down to th(» water. The 
place was sultrily hot and smi»lled disa<rn'eably of d<»cayintr plant>, so that we were 
very jrlad to ^et out. This was oik* of the few caves w<» visited in which tlu^re were 
no tishes. This cave pnictically joins Ashton No. 1, \\\r dividinj^ wall bein<r but a few 
vards wide. 

Ashton No. ^5, called Los Hanos, differs fr<mi the others in that consideralilv 


more of the r<M)f of the orij^inal chamber remains standinj; and t\w amount of 
territory covered by the water is also consi<lerai»ly lari^er. The* depth of the water 
differs from a fi»w inches to "overhead." This cav(» is occasionallv used as a Inith. 
At the tiuh* of our visit the water was covered with the usual crust of lim<». 

( ajio, •lai<riiim, and La Tninquilidad were visited by Mr. Hiddh*. 

Cajio, ♦*> or s mih's southea-^t of La (luira de Mcdena, ditfers <'onsiderably in one 
l>articular from all tin* oth«M- <aves. The wattM* lies at a nuich ^n^ater distjince from 
the entnince* than in the other>. The entnince* is an ordinarv sink-hole lo to 12 feet 
dee[). Inste^ad of lindin<r tht» water at the* bottom of this sink, as in the* Mcnlesta 
tyi>e, it is peM-fectly dry. L(Midin<r from this, however, there is a dark, narrow 
pjissa<reway, loo feet l(»n«r. which lead-^ to a very lar<f(» chambe^r with a cres(MMit-sha|XHl 
)>ody of wat(»r. This lon<r channel is not an inclined plane, but runs parallel to the 


surface. The thin roof of the nuiin chamber ha8 a hole blasted into it, throuj^h 
which light reaches the water. The fl<K)r is of falhui nM'k, as iti other caves 

The cave of tJaiguan, 2 or S inili^s oust of Cajio, is a chanil>er 100 foot long by 
40 or 50 feet wide. It probal^ly contains more water than any other cav(^ visiU^d, 
with th<^ possible exception of Pedn^gales. There is an ojKMiing in the c<Miti»r of the 
roof atl'ording considerable light. The entmnce is a small sink -hole at the edge of 
the cave, which is provided with steps cut in the rock. The* roots of a tree reacli 
from the central opening down to the water, a distance of is feet. The rootlets, 
very abundant here, were found to shelter many of the young blind iishes. Tn this 
cave is a central, rocky islet, formerly completely surrounded by water, as in Modesta 
No. 1. A narrow bridge of rocks has been built, which unites it w ith the entrance 
to the cave. 

La Tmnquilidad No. 1, in the Canas region, »^ miles west of Ashton, is of the 
Modesta type. It is a large cave, entennl only by means of a rope from an opening 
in the roof 4 feet in diamet(»r. Th<* water h(MH> presents an extensive surface, but 
nowhere is it deep. Large s])ecimens were tnken here. Side channels allow one io 
follow th(» water farther in this cjive than in any other visited. The greater part of 
the cave is verv dark. 

La Trancpiilidad No. 2 is a small cave with umvU light. It is in the center of 
a large sink prc)})ably (JOi) feet in diameter. The water is not deep and is easily 
accessible to cattle and swine. The bottom is formed of \ovy soft, de(»p nuid. 

The number and species of tish(\s tak(»n in these rnyos are listed under the head 
of the various speci(»s. In many of the cases 2 blind crustaceans, both of them new 
siwcies, were found to be abundant. One of them is a very gniceful Paht^immth's^ 
the other, a (■imlntut^ is nmch more abundant and forms a large pirt of the food of 
the blind tishes. " 

Mv attention wascalhnl })v Messrs. Wolfe and Ilmnunond to the IVdr<»irales caves 
near Alacnines in Matanzas Province, kV) miK»s east of the <»astermtiost of th(» caves 
from which Po(\v recorded blind iish(»s. Th(\v were visitiMl more* for the sake of 
visiting all the caves that might iM)ssil)ly contain blind Hshes than with the expc^c- 
tation of finding any. When near these caves we inquired whether any of them 
contained Iishes and w^ere told '"yes, but th<\v don't amount tx) anything; they d<m't 
have any eyes."" After this remark we folt thoroughly comfortable in a place where 
certuiidy nothing cdse contributed to comfort. 

Pedregales cave, about '5 miles from Alacranes, differs in some res[)ects from th(» 
caves of the Cafias region. The cnve slo^x^s down as st(»ej)ly as can convenient! v b<^ 
descen(l<Hl from a narrow opening, lait once inside it wiih'ns out, descending contiim- 
ally. The iloor wits fonned of a section of a cone that recalls the centnil mass<\s in 
the bell-shaiKHl caves of Onas. The floor and the roof were elaborat<^ly deconited 
with stulagmites and stalactit(»s, some of them united and ninging from 8 feet in diame- 
ter to a f Inaction of an inch. These* were pure whit<» when brokcMi, but tinted a red on 
the outside b}' tin* coral earth. When struck th(\v gave* out a clear bell-like tone, and 
the striking of various sized colunms by ditl'en»nt members of the party produced a 
pleasing chime-like effect. Among the caves that 1 have visited this is approached 
in clal)oniten ess of decoration onlv bv the cave of the fairies in Colonido. 

♦it-kH* W. B. Hay, in V. S. Nnt. Min.. xxvi. pp. 4l>iM:jr». 


Ai u vortical depth, jii(l|rod by the depth of a near-hy well to U*. al)out 75 feet, 
water was eneouiitered in the fonn of the usual crescent. There wjt'^ no indication 
that we had reached the l>ottoni of the cave and it is not known how dcop the water 
is, for roof and floor continue to slope down with th<* sjiine incline and stalatfuiit^^s 
rise from a depth of at h^ast '^ feet }x»neath the level of the water at th(* time of our 
visit, when it was alKUit 8 f(»et inflow its maxinuun heijrht. Tin* water was covered 
with a crust of lime and no fishes were seen. 

An account of the lVdre<rales caves will soon Im» issu<'d hv Mr. Pjtscual F<»rn»iro, 
of the CulMin niilway jx)stal service, a meml)er of the international copyritfht <*oni- 

Al>out half a mile l>eyond Petlrefrales is the M cave, so called from the M-sha]>ed 
])ath that leads from the surface to the water. The descent in this cave is mu<-h 
steelier than in Pc^dretnil^?^ Ji"d the stalaj'-mitic decorations nuu*h less elalionite. A 
dim lig'ht penetrates to the water. Here, as in Pedreorales, the cave fl(K)r continues 
to descend for an unknown distiince below the level of the water. Fishes were uiore 
abundant here than in any other cave visited. They were all of one sjx^cies. 

We visited another cave in Matanzas Province, al>out 2o miles north of the 
Pedrejniles, at the ed^e of the city of Matanzas. It was essentially like the M <-:ive, 
but contiiined no fishes. 


l^tween the western end of the island and Union, south of Matan/jis, a numljcr 
of streams run by independent courses from the watershed to the s<'a. Those west 
of Sjin C'ristoind are j^erennial. Those inuuediately east of San Cristol>al consist<»d at 
the time of our visit of a series of indeixMident j)ools. East of Artemisti the streams 
run alx)ve t^round only part of their course*, then enter caves and contimie their 
e<^)urse to the sea underground, or reapjx^ar as ''ojos de a^ua"* a short distance from 
the ocean. In the western part of this rc^trion, from Canas to at least (luini de 
Melena, there are no drv l)eds or other surface indications of drainaire. Farther to 
the west dry l)eds of streams, narrow and crooked, were seen, but with one exception 
there is noix^rennial stream lK^twe<»n San Cristolnil and Tnion excej)t alon«r the coast. 
The one exception is the str<*am formed by th<» larjre sprin<rs near (i nines. P^ist of 
Union we did not ^o. 

The Rio Sabanalamar at San Cristobal is alK)ut :in foet wide and varied from I foot 
to 2 inches in depth in cross-section. The water i> in j)laces very swift and shallow; 
in others "over head" and in j>ools. The banks of the river are clay: the )K>ttom is 
f^ntss-j^rown except in the deep ]X)ols and over riffles. We s<4ned up and <lown from 
the railroad bridjfe and also in an old channel of th(» river containin<if a mud<ly |)ool 
entirelv <'ut off from the river. The water of the river was clear and at IH u. m. had 
a temjM^rature of :^8 ( '. 

The Kio Pala<-ios at I^os Palacios varied from 5 to 4o f(»et in width. The water 
was (»lear. the bottom alternatedy <^ravel, nuid, and weed-<»'rown. TemiKMiiture of 
the water 23 C. We seined up jin<l down the ford at the end of the main street. 
R^tween Los Palacios and Pjtso l^eal the country is in part swampy, with lily ponds. 

The Kio Siin Die^o at Pjiso Ueal de San I)i<»cro is 15 to 40 feet wide with steep 
hanks alx)ut 20 feet high. The water was clear, in pools and rilfles, and 23- C. We 


seined near the niilroad ])ridjje. There are sovenil dry Ijed.s of trilmtaries in the 
neighborhood. At llerradura a small creek, 5 feet across and with banks 20 feet deep, 
was crossed. No tishing was attempted. 

Between Ljls Ovas and Goljx? were jwnds with white* water-lilies. No fishing 
was attempted here. 

The Kio del Pinar is, at the town of the sjime name, a broad shallow stream with 
alternating riffles, ix)ols, and w(»e(ly patches. Collc^ctions were made al)ove and 
below the wagon road near th(* ice factory. 

The Rio Cuyaguateje is the most western river of any size that flows to the 
south. Collections were made near Siimidero. The river near this town tunneln 
twice through rocky walls sevenil hundn^d feet high, and in one insbmce probably 
not much more than 200 feet thick, and in the other j)robably sevenil times as thick. 

The Almendares River is a de(»p and swift stnMim about 40 feet wide emptying 
into the ocean near Ilabana. It was scarcely ix)ssil)le to seine in the river itself. 
Collections were made al)Ove and below a dam at Calabjizar and in a small tributjiry 
just above the dam at Calalmzar. The tempemture not taken at the time of 
seining. A few days later it was 25 at Vento. 

The Ariguanabo is of special interest, inasmuch as it is one of the rivers that 
disappears in a cave. O^lh^ctions w(n-e made just alK)ve its (»ntnince to the cuve. It 
is a clear, swift stream running through the town of San Antonio de los Haiios. 
Above the town a dam at the ice factory has d(»e|H*ned th(» wat<»r so that a small 
steamer can run uj) to Laguna Ariguanabo. At the timt* of our visit the river was 
for a long distjince above the town l)locke(l with water hyacinths and other water 
plants. A much smaller though similar brook which runs through (xuanajay also 
disapi)ears in the ground several miles below the town. No attiMnpt was made at 

Collections were also made in the outU^t of the Yunuiri at Matanzas, l)ut we did 
not succeed in ascending to fresh wat(»r, and no fresh-water fish(\s were secured. 

Weascend(Hl the San Juan River from Matanzas to the h<»ad of tide water, where 
a shallow ford occurs. Collections were mad<^ in the ford, al)ove and Im»1ow the ford, 
and in pools of spring water. Inunediati»ly aJnive the ford the surface* of the stream 
was covered with water hyacinths, and the stream was 4 and more feet deep. At 
the ford the water had a maximum depth of about 18 inches, and in places formed 
shallow riffles. Helow the ford the ])anks Imm^ouic stec^p and the water is too deep 
for a collecting seine. 


There are recorded in the present pajwr 8(i s|)ecies and subspecies. These Ix^long 
to 25 genera and l^M'amilies. A numl)er of other species have been tjiken in the 
same region, not'ibly Lepff<(tsfrns trliitccohua. Of the 87 species and subs})ecies (in<»lud- 
ing the last-named species) l)ut 4, aside from members of the Githlhhv^ are found in 
fresh water elsewhere. They are the s]:)ecies of LepixoHieHH^ SymhranchnH^ Aijomni- 
tomuH^ and AiKjnUhi. LrjHstfstrnti trlxtarJuui is found in the fresh waters of Mexi(»o 
and the southcM-n United States. Si/ufhrmwhux iftfn'wfmrfus is genemlly distributed 
through the f resli waters of the tropics of Americii, Atujuilhi vlirytiijjm is also found 


in the .streuniis of oa^jteni North America, and AijinufxtonuiH immt'n^ohi is found in the 
fresh waters of the West Indies and Mexieo. Of tlie remaining si)eeies only those 
of the geinis Ilcros })elong to a strictly' fresh-water family. The genus llcm^ is gen- 
erally distributed in South and Centml American waters, one of its mem})ers entering 
the United States. The members of the marine family of (hUpikht are found in the 
streams and brackish water of tropical America generally, and their presence in 
Cuba is not so significant as their absence would be. A numl)er of the s[)ecies 
enumerated are marine, and their presence in the rivers may 1k3 looked ui)on as 
purely fortuitous: these are Tarjtou nthmtlnis, ]JorfjrhiiinjfJinxntttatfi.s^ Centropfjiiuim 

iKtU'tmoma^ and L(pphngt»ft!ii.s cypruinUhs. 

Two species, in many ways the most interesting tishes found in the region exam- 
ined, are meml)ers of the dee[)-sea family Brotulalct; the}' are the blind-fishes Stygi- 
cohi dt-niatux and Lucffufjasuhtcn'nncH'^, These have evidently worked their way up 
the underground streams and are now becoming readapted to the light in the upper 
courses of the streams. No other members of the familv are found in fresh water 
an3'where. Atlwrina is a marine genus with the jKiculiar Culmn s|>ecies as its sole 
fresh-water represenUitive. The remaining species are all memljers of the PaciUuli^^ 
a family inhabiting Vjrackish water and coiistwise streams. Of the Pwcllildv6 2 
genera, Glrnrdu)iuH and To,cti,s^ are peculiar to Cuba. 

The origin of the Cul)an fauna is then not far to seek. We have, as mentioned 
a))ove, a numl>er of marine s^^ecies, more or less regular visitors of the fresh water. 
We have sjx^cies widely distributed in the brackish water and coastwise streams 
whose presence is predicable (^rV/////V/,>), and we have local modifications of families 
with a wide distribution in the bnickish and fresh waters of the tropics of America 
{PdcllUdit), We have, furthermore, local adaj)tations of marine species to fresh 
water (BrofuUdi*^ and Athf-rina). The origin of all the alK>ve i> simple of explana- 
tion. The s^wcies whose i)resence is of greatest interest are the strictly fresh-water 
species of Lrpisn.sttii'S^ i'vicU'iitly belonging to the North American fauna, and Sym- 
hriitic'hni< and IL rfn< as evidently members of the South American fauna. The pres- 
ence of the eel in the fresh waters of Cuba is to l>e expected, inasnmch as it very 
prolwibly breeds in the ocean near Cuba. The presence of Symhnuirhu.s^ ILros^ and 
Lf'piiiOiiteK^i trista'<'hH.s and A(jtfnus(onui mt/ntlroht shows that the fn'sh-water fauna of 
Culm has a greater afiinity for that of Mexico than for that of Florida, and that these 
forms proliably reached Cuba by way of Yucatan. 



Symbranchus xnarmoratus r*l<K'h. T. I'iiiur del Kio. 

An^uilla chrysypa IUi(iiu.'»(|iie. Sitii Jiiaii, iii'ur lU* first fonl; Tumo Iv«*uI. 


Tarpon atlanticus ((*iivii*r i^ VuU'ii('ii*iiiu*H). M; 

IMiiar <U*1 Kio, 4 siH^'iinciiH, 20, 111), 1K2, and 11*2 iiiui., inmi u i1(m.'|> ihhiI U-iicatii tht* wa^on 
hridgo, iiiaiiy miles fpnii tliv S4si. T1k*v are ItK'ally known as "sadina," and we had Ihimi told that 
we Hhould liud them in this s]K)t. 


The menilx.'rs of this family an* every when* abundant, e«|»e<'ially in stri*anis iMinlerinjr the cave 

Kry Ut th*' (jriu'ni uj i\itmn Ptrrifiitl:* . 

a. Aiiiil fill (if iuhIl* Miiiilitr to that of ffimik*. i>vi|iun)Us: iiiti-Mtiiial raiml short, litth* coiivoliiUil: tiflh littk' 

movable; dciitary Uiiu'> tiriiily t'onnwti'*!: lower jaw stnniK hiuI usually {irojtM'tiiiK In-yoncl 


h. Tot'th all iioiiitci], in villlforin Imnds. 
r. Air hhulfU'r wi'll (U'Vi'1o|m'<I. no caiKlal «mt11iis; Kill-o|K'iiiiiK> ii<»t rf.slrict<Hl alNivi*. o|K'n"U".s fni* fmiii 

shoiildiT Kirdlc; <1i)rsal and anal noarly (r<iual: ori(;iiiftf ilorMtl in advaiicf of anal Funituhii^. 

('. Air-bladder wantiiiKl n blark iNflliis at r(»ot of caudal hi nialc; dorsal smaller than anal. itN origin Indiind 

that of anal lUvtUiut, 

h. Teeth tricuspid in one mw, no villlform Ijaiid of lei^lh; iNMly Hhort and iU*e|». eompressefl: dorsal «<hort. of 

H) to 12 ray-, fIrHt ray slender and rudimentary rtfprimntnn 

fvi. Anal fin in nnile idaeeil well forward and ni(Nlifie<I into a sword-shai>ed intnnuittent orKan. 

t1. Intestinal eaiial short. little e(»nvolnte<I; teeth all iMiinted. in luinils: eye normal: jaws not imMhieed; 

dorsal short, of (> to 10 rays, iN'hind origin of unal: mouth wide. I'hin \o\v HniHUHi^hi. 

tid. Intt*stiiial canal elonKatt*, nnich convolute<l. 
c. Teeth iMini|>n*w<l, entire, without latenil euH(is. 
/. Anal prtK-esK in male very lonj;, Herrate U'hind near tip and with fhiKer-like claKiier» f^the pn-puee 
beiUK m(Klilk><1 into u ]»uir (»f cliiN|is); dorsal in iNith sexes Inddnd origin of anal. 
;/. iH'ntaries and intcnuuxillurics firmly united; ti*i'th i»f outer n>w tixe<l. a Imnd of minute tL*eth 
liehind them. 
h. Teeth of outer n>w much eximnded at tip, bntttdly s|MMle-sha|ied in up|H-r jaw. eIose-s(>t. tlieir 
nuin^ins overlappiUK; teeth near middle of lower jaw Hsymnietrically ex|Ninde<I, latenil IoIk's 

prolonged and ending in n |M»int tibtrvhiMhiti-. 

hh. Tei'lli of outer row wide-JH-i, si-an-ely exitandinl, s|H'ar shapitl. those near middle of lower jaw 

in tw«> irregular si*ric> Ttwui^, '> iu»v 

I/.'/. r>enlarics and interiiiaxillarh's I(M»sely joiiieil. Itrlh of outer n»w movable, Inseftetl <»n lijis, a 
few teeth ludiind them or none, llnwt.' of outer row wiile-M-t. M*an'ely exiMindi'd, s|iear-slm|K'd; 

tho-e of middle of lower jaw in two irreKnlar series Uirnrtlitnih. 

ff. Anal prtK-»'v«. comiuirntivcly sln»rl. a Icaf-shaiKMl prepuce attache«l to the anterior surface covers 
the tip. lip without clas|»«-rs; diirsal in female in advance oi ohuin (»f anal. 
I. Tip i»f anal pnM'«-ss in male ending in a simple antrorsc htNtk, \\u serm- on ii? (Misterior Mirfa<-e. ..hirilin 
fi. Teeth all |Kiiiited. origin i»f ilorsal behind that of anal JhOnntilun. 

Fuudulus cubensis Ki^*nmann, new s|Kries. * 

1 am Honiewhat in doubt as to the generic ]M>sition of this s|H>eies. its short intestine, double mw 
of teeth, unrest rict<Ml )rill-oiK>nin};s, an<l ]>osition of its dorsal in relation to its anal, ami similarity <if 
sexes (at least one of the thn-*' siHM-imens is a male I Mi-m to iinlieate that it is a siMiries of Futulultu<. 

TyjM*: No. iHMiT, Ind. Univ. Mus., L^J mm. lonir; I'inardel Kio. 

Cotyi^^s: Two s|K.*eimeiiJj, 20j mm. lonjr; IMnar del Kio, at the fonl just aUive wa^>n )»ri<ljre. 

Head AAiy alMiut injual to depth; I>. 1 1 or 12; A. 10 or i, \i); nales 24. Orijrin of dorsid very sli^jhtly 
nearer tip of snout than l«i.«e of middle caudal rays and over the ei^fhth scale of lateral line; orijrin 

"The following «-harucicrs an* used to indicati: the Kenenil distriliution of the Kciieru and N|»i'eie8 eiiumenite<l: 
t, i:enii> peculiar to (.'ulm; ♦. s|iecics |»i><uliar to Cuba; T, guueruUy distributetl in trupicul fresh watcn; M, muriuv bpecici". 
f' To(ov»u quiver full of arrows. 


■>( aoal uniliT (-levL>Dl)i soilc. Ihjrsil anil aiutl liiuh, miliar l>«.-hinil, lar^t ray Riniewhiit proiluceU, 4.3 

ill U'ti^^tii tit loH:- of tau d. a 111 ra n-a g la.'H. ii la iwi la nearlyso: 

ventral^ rtsichiiiii Id en mraa orau a and wry 

obliijiii.-; eye lun^T tn n-u i«a nUiru ^rhn (runt, 

slightly rauvex towar r-a T« ii ta a., » sin« >-t. fou -t es ilanftd- 

Cokiratinn britlia du -uuauu n.n -v-Kt'tiiv of 

spite; a ilark liunif ra a r^ l>aal>>a r<t!i lower 

jart of fve fiirwanl K h a L a i i-iijics, 

lnmicriHl hImivi- aixl I n u p Ik-Iow 

origin <ifi|.inal; iMi-k a k n h ucr lifilil 

banilu'lu-'ky liatKlu an V akn n axil of 

l>«T:toral t'lliiil; l»'lij» jr m •« ni anal; 

U-l'.w till:- i:: :i narrow m k a. » rl ipii.i-nt- 

itsa luiiil; i-'W n[ I, i,la<-k; 

vi^iilral surfaco ollnT« w u 

OambuBia punctata I'"!- 

Sin fristolKil. --HJ (vmaU*:;! I.. 

S4 lUKi.: 11 

inak-^riOt'ilil itiiii.; )V-«> K<-al. 

■J (Hiial.-s 

Ri... !l fiiiiak*, III.- brvi-si M nil 

II.; San .In: 

llif lar)n'!'t 4:; Mini.; Calahiuuir. ; 

r t.-nialfs. 1 

This -^i-i-i.^ i.-' ^iK.rl.-r .i[i< 

1 ,1.V,,T .1 

■lorsal in ilii- s|.,-.iiii.-n>: tiikeii <■■ 

Slid;-, til'' 


It, atiii II); I'ulat'ior:. 1 ft-iuuk- :;!• 

mm., ly.i: 

llir larm^t i'> [iil[i. 

lilt 11. as (iarman I'liiiiil in hi^ >lifciuii;ns, 

, l>. ^<: Siiii OistoUil. 4 ji'uiali'^, tlu' laiw-^l 4.'' nini-. I>. K 

nar .1,1 l;i... 4 fi-uiali-!-. tin- larv-iM 47 mm., iiml .1 inak>, 


liiti'stiuu) lunal t-loiipili-; miili-?^ uilli una! liii nit>li1ii-<) inl'i a very Inii); intromit ten I oruan; ja«>< 
iniii'li more lirinly iinitiil than in '■■ftr'Unii--', eaili with a mtii?' of eliwe-^et. Iifiail-tii'iifl. entire ur 
but sli^litly ireniilale tii'lli. IIicm- livtii not miivalile. a n^irrnw liaml nt smitller. l>roai]-lip|ie<l. I'oiiical 
orlriruxpiil teetli N-liinil lln-iii; tins >mall. atail in a'hami' <>f <l<>»al in l-<tli h-ncs. 

Mor-t nwirly aliii-l to i;...-l..< aw\ i;.r-r,l...ii^. jiti.l them ineharailer of l.i-lh, '.■—/... 
Iuivin):tri|-ll!ilii<l leelli, i;ifir<lii."i- haviii;" tnnvahle li-K-ely-K't, and Tiuiw liavinj; lla^■tat^■ teetli. 
Ol&ridicliUiys umnotatuB I'ih'v. * 

Alalii'lanl. Klfly i-[H-.imens |.rx-ervt.-i from San Cristolial, 47 to ,S4 mm. hiny; a niimli.<r of 
i<|ieeiliienK liave, in a<l<liti<>n to the lalenil s|.>l, a s|iol on either Mile of ami^ ami MfUiutime.^ a Mack 
streak eonrieeliiig llu- two; in one iui-Uuce tile lateral fjiot on out >iile ie entirtly ivpLieeil l>y tlieanal 
v|iul: niakv 1 13) :t8 to 47 mm. 


At PaiocioB 2i Hpcciiiieii 
in Stui CriHtolml Hixi'iiiit'i 
I'bmi Hull, 2 fciimk-H-, 1' 
liuiur wrieM td teetli in iip|«T j«w kii 

■T JHw litllit expujidt^i 


oKl I 

I Diie H[H.-eii]ii>n iinly 

ilcl Ri«, 4 fciiial«>. 

itli fharjily triuii^uli 
I. Ti-etli of oulttr n 

II. FlucliiHliuiiH iu the lut4.TalH|iot8o[ female 
tiall H|>ot i>n iiTic nth', iioiif on iilhl'r. 

T trkiirii)i(i ttvtli ill 4 iir 5 rowa; 
ii<l<lli-i)f jawa irniiulurly 

t'xiianiluii, lateral loln-siiroloiip^^liiitoa jmiiit; latum 1 ti'otli of lower aiiit of upjM'rjuweijuiillyexiiandc'cl 
01aridiclithy« falcatus Ki)!i!niiiutJti. new Hiweiex. * 

Type, K<>. fitHM, Iiid. Iniv. >Iuf., a [t^iiiale, 82 luin, Itniti. tn>in San CriHtiilial. 

Cdtyjies; Ki(;ht feiiialta fnaii aui.ild rivfrehuniiel at San (.'riHti>l«ii, tlieHitiiilleHt (iO inni. Icmft, (lie 
lar^wt ttT) nini.; 4 feiiiiiUi) 5(1 to 5it mm., from ralat'iof, taken in a niiidily (miil in tlic river liLil at the 
fiinl; 8 femali-B ami 2 niaU-s from Kiii del Ilnar, tin! fenutlui 118 tti 47 mm., tlie malef 2it and .17 mm. 
Tliis 8i>ei^iew rciK^lieH itB maxinmm Hiie and i« most ubnndaiit in wann. muddy \xh,U. 

lirmtl'/-lUia«r EljWllDl 

Iloily liHif!, aleniler, Jittk" eijni[m-»«<l; luail 4; deiilh i ',iii ]in-Knaii' feinakf 'i.^); ]>. li; A. 11; 
M-aln' 2!l; lieiul liroiiil, weiljje-nhaiieil in iinilile, with lowfr jaw very rihlii|ue, jirnjii'lin);; eyti very 
litrge, l(in(!<T than muciiiI, in hea<l, 1.4 in internrhital; mouth very oltli<|ue, wuall; inlerorbital 
iliviileil inlo II difitinet riwionn liy km^itudinal icnxivcH, ecnlral jxirtion eiinvex; ori^flii iif ilorsal (!i)ui- 
iliHtaut fnim \mM- of middle eauilal ruy anil orijiiu of iMt^lonil; dDnul and aiial falcatt'i xei-oiid ravK 
Hiekle-Nha), ctach t'xIendinK for one-tliinl itt* lotul helclit U'^-iind tip of laft ruy when foldiil liack, 
little lew than It^utfth uf head; caudal emm^nmite. wmie of onter niyn prokint!e<l; origin of amil in 
fcnmlt! alKiut ctiuiiliHtuJit from liato: of middle eaiiclal ray and anli-rior inaiyiii of eye, iti< seventh my 
under oritcin of domul; ventrala uxually n.'«i^hin)c to anal; |icvturalH alKiiit Iu middle of ventiali* in 
ff'Tuidi', to \tami uf anal in iiutle. 

San Cri8twl>al ijpeeiineiui very jiale; a rliiKky Htreuk from nu]>e ahmii middle of Inu-k toraiidal, 


.- lali-r 

>iiil-lt' x-ric- of . 1m>riiiit.>|.li. 

r;ila.'i.K-,-]«-.iim-nr-...lur,"l lik.- tli..«. ir..ii, S:iii Crir-ii.l«il..-xi',.|,t nm-iii whicli ea.'li w-iik- nf tliesi.k- 
U-l.iw Uti-ml liri.- i> .-k-i''! willi a ^-ri.-^ .li rlir..iii;.lii|.h..rt- ami ihtri- i.- a (aiut liiiil nf K .lark Hj«tt!' 
uloriK' ii><-<iiaii Mark litir; rvvioii uU.vr 1iit.-r..l lltic .liL-ky. 

1-iiiur -M Ki- .'ixviijj.-n.' ,:.\..r,-\ [\kv \hi- 'Uirkvr l^ll:>•'in^ ^|K->'iijj.'ti. s.>LiR'tiiii<-.- :i Murk strtrak uii 

■■itln-rw.l(-..i allU^a]Jl^ forii:iri1 r.. viutraU, Mai.- "illi Iili>-.| [".rti..(i i.f tlic anal vi-rv ]..iii;. »ilh 

n.-tr.irM> ^]-m-> l--liiii.i aii.i a lit;l,- . liu-|-T ..ji ii[. ..i Imii-,..i ray- 
Olaridichthys torralbasi Miu'inniiiini. ii.'w >|«'ij.':-, * 

Ty[«', N... •■Kr,-2. l„.\. rniv. Mil. ,'. r< mo.. |..iii:, imn, l'i„ar .1,-1 Hin, r.-j.ix-M-nl^ a|.|«ir(-ntlv a 
IR-»-s,K-.-i^.. Ill i;i-N.-nil :ii.i.,«r-in.-.. il >.rv L-n-allv r,-H-uiM.-^ III.- tll:ll.■^^ ..f '/iV-.n /;».<:. m.l.-llirua ail-};. •n.^..:U,l..,. Kr..iii tli.- f.iriii. r it .liff.-is Jii [".-.-...m,,;. lu,,,). ..f |,.,.th in jawa U-)iin.l 

fnm. Ilu- ^-.-.iiiil il .iitfiT. ill .■n|..nili.iri. haiiiii: tm ^j-.t uml !i .■..ii:<| .|.,rail Iwiid. There 
un- ii.> <>tli.<r ^■jH-.'it's wtiidi il,:. 

i>. il-. A, HI; r-ali's L'K; hfiiil ;i.ii; ■ii'i^lli :i.:i: l.-iv rM[i,|irL-M<-.i. .-l.tiijniu-: im-iiUi siiuill. ^ulrti-niiiiial, 

.iiitiT «Ti,-- iivi-rlaiJi.iiii;, llnr-i- -i l<i»i r jaw iiinri' ]"iiiik-.i; :i liaii.l «i iiiiiiiiK' Uflh licliiixl at itabt 
till' (r.uit riiH ill iiiubt jiiw; 'lor>ai Muall. il,- li[^t ray i-.)iii.|i-Unit ir.nii Uim- •>{ iiiirMl,- c-judul rave 

tyc: cailJal 
loti;:. niili »>rni- U-liiiul, ami 
yi-; {Hxtiinil n-acliin^ lo aiuil: 

<>ai uilh an'l liaii.l r. a. liiiiL' fn.m t'liH l.-iiL'tli <>f lin-1 ray l.i Uis.' ..i \ar\. i-anilal, vi-iilral. aii.l 
|*<i't..ral .-..lorl.-.-v-.; liivl ray <>t anal pnK-.'>^ laiv<-ly 1.1a. k: r^i'l.-:^* witli al-iil in .lark .-n,».yln'iiks: all 
s.■al^■^:! Hilli, <l..r^al nii.-i.ol li.'avilv ^.; a .lark .l..r>^il :i vi-iitml \\w 
lN.-liiii.l atuil. a .lii>kv loiii.l an.ini.l li.'a.l jii-1 in Imt.l ,.i vy.->. 

I uik,. i.l.-aMirr il. naiiiinv this -r-vi,..- ...r I'n.l. ,I.,h 1. T..rrall«-, ..i Ih.- .hair ..i /.-.l..w'V in tlu- 



Oirardinus metallicui 

■alt- I. 

:i..; lariir^i i„al.-. 41 

-. V> n.n.. A f,-H fi-ii.a!i 
r.mi I'hiii t'. tail lila. k i 

Sai. An1.>]ii.>: lar^.f^t ft'inak. ?i 

Ar^hlLii; f.'uiuk> usually »illi 1>la.'k ..n anal: laru. .-t t.-inal.-. -'<1 1..111.: latvi-^t niak-, :W nun. 
I'inar ilcl Ki.i: larin*! fi-niak-. tl« tniii.. Iari:it"t nuili-. 41 nini.: a nuinlK-r .1/ uiai>» with a lilai-k 
struak iif varyiiiB iiituiieity ami wiJtli aluni: llu' vi-nlnil {.iirfaiv. 


Olr&rdmus gumani Kifn'niiiaiiii, lu-w Hpt«iif. t * 

Ty[N', Ku. HHI>\, Inil. I'liiv. Mtu., iiiic iiiulis :V\ riiiu., I'iniir ilcl Ri<>. 0>ly|ifs, imt' male, :S!t mm.. 
Pinar tliil Kii.; .nu- ^Ilal^^ :« iiiiii., l'Hliu-i<>><. 

D. 'J; A. !l; Mvit«< 211; <l(>i>lh .t.-l to huiul :i.K lu :t.S. IIihIv <i<iLi|>n-xM-il; lit-iiil tniiicaU-*!, luwcr 
jaw iiottrlv VLTtiml; miiutli very MUHlt. li]iH llii<:k, Ut'tli in ii fUisli- itTii-H in I'lU-li juh-, hIi>|n!i1 ua iu 6'. 
iiieluUii-wi, very imivHhlt;, inU;niiuxitlHriit< ami di'iilork-K tint iinitcil: i-yi- ii;' I'ltij; uh fuimt, 2.^ iu liewl, 
uiual t« iiiti'nirbihil; oriKiii iif ilorsnl u lilUc iii-ariT iKi'turul tlutii lnkM- i)f miilitli' laiuilul royti; iliinal 
roundel, Hiimll: tloncil, inurlal, anil ]>c«:Iiirul uf uIhiuI (njiuiI liciirlit, ii|iui1 tci ilii^luiKi- iif [Kt-tonil fmin 
isyv; aiiul |inKvw 2.5 iu luUK'th, svrniti^ ui-ur jtH lip lH>liinil ami willj it rla-tjicr iit it^i cucl; votitralK 
very Miiiull, ruu'liiiij; to thi- uiwl; vi'iilral [<iirfaiiM-olctrUw vsiviil ii lilai-k line frmn iiuul tti iiiuilul; 
Biuliti of Hiilu with a dark margin of ini'TvuMiiii; wiilDi towuni tuu-k: » dii!>ky ilonwl stn«k: Iicud in 
front of eye 'lark almvu uiirl In-low; irniou Ik-Ihw i-yi- i-iili'rlfi<:i; a ncll-iU'rMitxl ljliu-k sjmiI iiu liuw of 
ItiHt 5 dorsolravH; firut donuil niy lilack; uiial iimcL-XH IilLU'kixli >iii ImkiI liulf, u nmiill iiKliHtinct lila<;k 
ti|Hit on (liHtul liulf of liut anal nii^tid>mm' and i-xtt-mlii^; at tiuio' nn nvifcliUiriu)! n-);ioup; uidi-^ 
witiiimt Bln'aku or luire. 

TIhh ><]M.i;i('H iliffi-ry fniui '.'. itui'UtwHH aiiil O. ittHlii'iliiiui' 
Hixitti or lian! nu lliu fiitt.'M, in tin- iiumiIht of anal rayi 

TOXItS Figanmaan, aav ganoi.t 

TIjjTiM KiK-cimuinii, m-w t.fuu» of i'..ri/;,W;i (,„l,ll,'i). 

1\m >;.!nnH ilitf.rin Innu flhirUUi-hlhu' in itt- narrow I.fll., from i.'ir.,,,!.,,,,.. la liaviii^- ite jaws Hrmly 

Tozius riddlei Kii-tinmauii, ui^w mini'iiv. t* 

Tyi«-. N.I. «»!, Inil. I.'iiiv. Mu^., a. Uwah; :.! i. |..iin, fnmi S;.ti <ViH..Ud. 

C»ly|H.«: 1 funiak' (Mi trim. Ion;;, and :! mall's :i:nind ;ll ii.ui. lou^-, if.ui San Cri.-<lol>al. Hcwl 4; 
d(!i>tiiJ.4; I). II; A. 10; >Nul<3iiL>K; ori|;inofilomil midway Ixtwi-iii l.:is)-of niiddUcaudut r:i\iiandoriKi]i 
of upiKir jMrtorul ray anil owr thirttvutli m'hIc of liili-nil liin'; 'iri;.']!! oF uniit Ih'Iow i-lrvi-nth walv; fins 
moderate; lunt^ust d<ir8a1 ray niuaiin); lcu;;tli of lii-ud witinnit mihiiI; i'vi^ I'ljiiulin;; Huont. little mure 
tluin 3 in IimuI; iiitvriirliilal i-onri'X, iiiiialiu)! tinoiit and cyi-; |>riitili' piiKlitly curviNl; initvr row c»l 
tevtli movul>U>, PiHsir-aliafHii, nut vi^y d'xM-iy net, tin>wu-ti|>|, a Uuid of mliiiiti' In'tli liohiiid tliein; 
a <lark latfrul liauil crotued liy a)x)ut I! indiflini-t dark rruwiiaiiilr'; tfuileH of !iidi--> wilh u luht oent^ 
auilauarniwuroriirouilcr margin of dark, fiirmiutin-iiiiilati'Mis; ail)irk>>tr(i!(kfXli'mliiiKtlowulN)tWflen 
cyit and annio of mnutii; iwat of liiwt'r side of head and Urlly wliili'i )H-clom) tolurlcH*; caiulal (aintly 
diuky; aiml willi a faint dark liunil tliroii)tli the nddillc, the ti|>H and iiasi' (ijturtt'HHi dursal diiaky. 

Mall! muili xmallir, tin.- wior coutnu^tK cliari-cr; alHiul 7 wcli-markod dark rrotvlundu in tha 
lar^ur H|ii-i:iincn ; dorxui li|i|KHl with dui'ky, a black lianil from kiM-of l;u^t dori^d rayx forwanl lowwd 
bauul lliird of fourtli dunnil niy; in the lart'LT H|ii.-cimi'U it wrii-x of 'lurk s|H)tii on ilorval niya on a level 


will) (i|i lit lint one; u Iilui'k 
iBKt ravs; first (ullv ■levclojH 
iiiiul uixUt oit:lit1i h^iU-. 

•J iiial**, iIk- larvivt 71! im 
sliiiwiiii! uri'iLl vurialiilit; 

iiiiW. (Ill- larvi'^t 7<l]iii>i- 

PcBcilia vittata (iiiirliMi'it. * 

Sac. <.-risl..lwi. .>v,.ri(l U-„iak->. ,]„■ 
nioti'ht-K; (III- ^[tiiill ••111^^ Willi :t y<-1I..u ^tri|.i':^ 1«-l<>w III.' hit'nil liix- ui«l fK'.|(ii'iitly uitl 
sorics iif Maik ■loli' alunu I'lwcr (uirt nt si'lf; ^L' r 

Suu Aiituiiii'. 5 fi'iiiali-^, till- lar(,f»l llfl iimi. 

Iaih I-aliu-i.c, 71) fi-Miali-K, Hit lanrtft IIXi iim 
aliiiii; ;!5 iiialcs, tin- laiw' »>5 '■""■ 

Calalur^r, !) ft-iuuti-x, ttii- laiv>?<t Itri uiiii. : I tiiiik-i 

Plu^i Krai, :l fi-iiialix-, tlif laivi-^ SO mm.; I male. '>:i iiiiii. 

[•itiur .li-l Kill. 17 f.'iiiak-:-. llit- larv-wt Ki mm. : » nmU-s. t!ir hitvi-^I l« mm. 

Siii,ii<U'ni, -2 fi-mak*. 
Heterandria cubeneia Kit;i'iiiiiumi. iit'iv iiiH-t-itv. * 

Ty|.«-, N.i. TW;:!, lixl. Univ. Mli^.. a kmal.-, ri'iiiiri... (r-im l-alaii.*^. 

0)ly|itv, iini'd'muk- witli yniini:. *!tiiiii., (rifiii l.<is I'alarirK,, umloiif (vniali', :t.~luiiii., from I'inarik'I 
Hill. TlitvL- t<ptrimi-iii> a;:m; well willi lln- clianii-liTs nC ilii' miiu:- Ilihr-imlrUi, a» rei-lricti-il liy tiar- 
nian, t-x.vpr that tlit- oiittr xt.-rii'i' of t.ttli ait- itKivaMi- llt-ail 5 to r>.2: cl,-]>tli :!.:; t<j :;.7; D. !>; A. lU; 
w'ak's 21*. IkHly clnntratc, link i'ijiii|irf!i«.il, t-viitTjl aliajw ikal of t'tiii'liJiu-; |injtilt' nvnlarly filrv«L 
from 'lonml tin ti> cycii. Hattoiitil ifver eyt.-:* ami foruaril: moiiili siiiull, ojicitiiit; U|m'ar<t, the lower jaw 
|in>}(vliii)!: l^jiieaof jaw loosely tiiiitcil; I'vc 1ii[i«it tiian mi.mli,2..S to 2.7 in Iii'U'l, very liltk lei« than 
intenirbital; origin uf ilnrsa I a little iifarxThi-u'I (liaiiluiwnf iiiii|ilkniinlal rayi>, civerlirBt thinluf anah 
■lonal anil anal both lak^tv, the aiiterlur ravH exlemliiii; iiiiu^iilerahly beyoml ii|> of laf>t when laid 
liai-k; lii;;liest (lor><al my sli^flitly shorter than hijiliii't aiial ray, alHiiit ojual to lenj^lh of head; i-audal 
a little lotifnT than liead; veutral rvaohiiiK to anus; ini'lunil ruaeliin^ to ventral. 

Sealee of tlie niid-doreal line with their doraal ludvee diL^ky, thutte of entire xidv marked with 



black, inout diBtiiict alxivt lateral lini-; a wriw of 12 imrniw ilurk ■ 
aiid ae liiKh ox vyu; an irrtiKulur lilai'b slrfak iLliini: niiildlt' nf xiilt' 
niembrutii: lilai-k; a lilac k lim^aii'l a ffw diriiiiiatniilnirt*aliniir '■ 
to caudal; chin <hiHky; i«Tt(inil, ventral, uii'l ln^lly wilorkt*'. 

L-rtii^l Imnilti uUiiit um wide au pupil 
(liirnal li|i|>iil with ilunky. tlic tint 
i-h iLiial ray; u blai'k line from aiial 

■A ^\i 


DoryrtaamphuB linoatus (Val 

I lli<- Itio Knn .Tiiiiii Iwlnw ihi' fonl. 

Atharina evermaniii Ki^*nuiatiii, ii< 

Tjpe ^ fl657 I 1 U M 
■Ml fro Su ( Ht tial H 1 1 

1 I ilititai -e fro 

lenKtli of eye; caudal litllu \vat tliaii lenfifth of Ik;iu1; arm) immtIuI in lulviiiK-e of oritriii of dtinul; voti- 
trala aiiiall, imt reacliiuiiunal; (irctDnil rcacliiiijf ti[iH uf ventralt^; a eoiiPiiicuoiiM laterui Inuid iiiuxt 
inUiUBe oil cauilal {Niluiicle, ^raduaily fwliiij^ out umler jiectoral; n*;:!!)!! uliovc tliiH in all caMM thickly 
pL'jipcnxl with iilack celiH, niont thickly hi> nlniiff iu<Hlinii line; rvKiun U'luw tliiH in many cancH 
miiiiliirly but Ivw iiit«nHuly Hjiott4:>il; liiwur Hide of huail uiid lin'iiHl whitu; veiitralN iiinirly fret- from 
|iiKini4it; all the other dua with i>i)[iiit!iil, uvllu uf iireater or 1i9m inleiifity aloiifc the rayc. Ttiii> Hiwvii-M 
in rouilily dit!tiii)j:niitiiable from the other H]K?eitii iit Afliiriiiit by the mnaller iiunilxT of Hcakv, 

I lake ]>)<!:ixuru iu hauling tbiii HjieeicH for I)r. llurtun Warren Kvtirtnnnu, in reei>)!iiitioii of bin 
valuable work on the tixlieu of the Went luditw. e«pM'ially hiH work ou the liHhva of I'orto liicii. > 



AgonostOmus mouticola I )l:iiiiTi>Jt |. T. 

lann-ft 171) iiiiii. : I'itmr <W] Ki<i. n!iiiii<k[il. 1li<- lar.'i-^t lilO i. 


varyii.u' i nt'tl. ir..iii -m t.. I'.MJ ,um. 

laUtiauuB jocu (llliicli A S(-lin{-i>liT). M. 

IiUtianuB grUeus (Walliuiitii). M. 

ciiH- pdiall HjiwiiTieii (rotii du' l!io Siin .liiiiii, jus't U'l.iii tlic fonl. 


Eucmostomus meeki Eip.-niiiniiii. iii'«- ^|Hi*ii-;^. M. 

Tj-iH-, Xr,. Wm. Imi. l'niv-Mii.-;.,.is|«<;iiiir(il:!.->(iiru., fn.iiiSiui.Iililii Uiv.-r. jn.-H U-L.iv itw fin^ f..rd. 
I ieiienil ai'iM-aratiii- of f'ixuui l.fn.iii. liiHcrin;.' fnmi all utlitT cpniiin c.[ tin- ^tiiiis F.iinuiHii.iiiii/ \a 
having liut 2 anal !i[iiii(>^. 

UkuI ;{.i'>; ili-pfli ;j; P. I\, III; A. rL, S; wiik-s 4-l'i-!); cvc^ ! in ^n.iiil. :i in lu'.iil. I iti iiilrnirliital. 
IWly i-loHi.'HU-. littl.' <-.m.i>n-s^-<l or I'lcvaU-.!. ihi' .lorsxil |.ri>l'il.- hnt liltlt- itiorc <'l.'V!il.-.l tliaii the 
vcntnil; siLiiit iHiiiilol, Iti<- |>r..1il.- fri'in hxxK to il'.rKil ^I'uily!; inniitli narrow, It'rniiiial. I>ut 
Littif till' lowiT iiiarjiiiL nf tlir vyc; maxillary irachini; U. v>Tli(-al (r..rii trout of vy.; :t.4 in head, 

I'irli', .') ill lii'itil; itili-niiaxillary 
Ifn>i>v(-t'nlin'ly natnil, its wiillli 5 i[i tlic iiiliTorbital: iirfiiH'rcli- ami iiiroHiital entire; •lor^il R|iin«< 
Hlenitt^r, tlir pu'nnd ImipT^, 6 in Ilii- li'n;;th: vi'> slmrt, roacliint: lialf uay to atial; |Hi'toral lon^, 
3.5 in thelcnf;1h, rtwhinj: U'vimil li|)K <d tlii' vi'iiir.i)^, hut not to vent, t'irpt anal Hjiinc ininnte, the 
Hecond tHjiial to the k-ntrth of ihc < ye. 

Color, aHhy cray. with hiiiik' inftallir n-llo'Iioii)'; 'liixky lin<'» alon^' Ihi' t«wi' of m'uli-)'; xidrat am) 
hHi:k evt-rywhtTf puiirtalt' with ininiiU' lioti-; vertical tut^ •tni'ky; \i'Htnils and jxi'tdnili' li|-lit<>r. 

XaiiKfl for Dr. Setli Knceni' Meek, a»*ii4aiit curalor of zoiiJony, FiuliL Coluniliian MiiHeuni, in 
recognitioD uf his excellent work oa Mcxiuan fiebes. 


IiiiliviiluiilH uf the tnmiis Ilenii an- iih iiiiniiTiiiiFi in tlii' frivaiiiH iif ('ulm oh imlividiialH of the Oni- 
tmrrUidir are in tlic rtn-ftitiH iif iiiiuil wzi- in thi' CHiin valley. Thry were foiiiiil by iiixlown tn liile 
WHter, Irtit nut in if. Dniy a sinirli- Hiiecii-s hiu" bti'ii n-eonliil (nun (,'iiImi, ami nofhinn him l)e«ii wuil 
either mnntrninif iLi iliHlrilmtiiin <ir ilH varialinn. N" "iio. exiT|>l |K»iih!y I'oey, haw liefurp liiin I'oni- 
juinil niiinlK'ntiif HiHi'iiiienH f nirii iliffi'reiit |>)ai'o>< or even fmiii the »iii>e {ilnei', Surh a eixtipariwin 
in iiierefiin' very ili'riiralile. unil thi- iiiitlcrial r<il]c<'ti-<l (iir Miirpni'MO' nil olher rnliei^tiiins iiiade Iiefore. 
We have nltnjp'ther 2;ttt p])e<;iiiienH frmii variixm liN-aliliec. All cxuminati<in of all uf tlin*' proven 
either Uie priwini' .i( Hi-veral iiislejul of ji Ati]i\f wiH-eieii rm the ixlniut or n Tc'niarkalili' varialion wilh 
tiiealili<f, A clelinitimi <it the variatioiiti hw iinivfl very vhifive. The nunihcrH ot fin raycanil M-aii-t) 
are iiniforin, wi Hint the ■llfteretnti' exist ii) tile jiriii")!!!!.!!)! am) tiic ciilor. But the eiiloralion a\go 
liai^ a icrlHiii iimlerlyini; iinifiirniity. Thon- in a spot ni«r Uie niiiiiile uf the siile, Hnolher at the l>a»e 
<if tiio iiiiiiIhI. aii'l mi aliwuire Ihinl ulKive the ifill-ojieninii. There nr>' iiiimeroiiH Hriiall f\*ib' on Ihe 
rmt> ami on Hi-ali^ of 1hi> si'len, <-^|Ht'iHlty Ih-Ior- ami <m the 'i|M'n'le>'. ami winietiiiKw on the <-hK-ki>. 
Then' is al*i a liinirilii'liiml flrrak from Ihe eye thrcmgh the lateral H[>"it to the r-aii<lal sjHrt. ami a 

definite niiinU'r of i r'ii«il>iiri', iKitli ntreak ami Imry imist i'iin»>iiii-uoiii< in Ihe yunnj: iiiiil in lijrlil- 
colnroil iiihilt itKliviiliiHlH. Thin nniloniiity nl iimlerlyiiit; ftriK'tun- niakit" ileliniiifiof s]H<eie:<or varie- 
ti(« a 'lilhriilt iiriH-eeilin^. The ])<ilynii>r]'hiNhi i» fnrther <iiio)i1ic»teil l>y inxlanirH like the follciwinfi: 
The »i{ieoinienK from San Antiiiito an- muUly referalite Ui ii ii;rtHiii form touml at Calalmxar, although 

they (lifter from C'aialautr I'lH'eiii n in <|nlte renilily <lii<tinjiuishal)le featiirtv; )mt one of l.heni differx 

notably from all other H|)ceiiiiens mlh^eted at San Antonio, and wonM iiiihi'tiita1int>ly Ik- i'<inHi<h'reil n 
Npec'i(« (liHtinrt from the other i<|Hi'imeiiR Irom Ihe Kinie lixality. Itni iil fatattioK (he wiiDe form 
branehinf; from the Calaliazar fonn apimiai'hei'lhei'hanu'leri'of theHitigleHiHi'lrrien from San Aniouio, 

1 venture to tliweriiNt huru ei>rtain of the mont alwrRiiil formn aw new, withoni, however, fet-linjc 
Hiire that th(>y are really diBtinet varieticH "r njieeiitiorthal »oiiie of Ihe other forniti referral to //.Mm- 
con/Atu are not alwi new. 

HaroB tetracanthuB Cnvier A Valeneienniti. ' 
HeroB tatracantliUB torralbaai Kii^iimann, iii-iv i<n1it>i>erle»i, 
(2.^ Hp<!eimunR, lit) to 181 mm. Ions, from t'alaliaiuir. ) 

Thctv^ F>|)ei'imenH romo from the Alnicnilar(»> Itiver, 
prol)ahlethattlietyi)eof llrrm Iftrnamthun rameflom thef 
I'oey's ilrawin);, on whicli their trlmmnthwi wbh IibmhI, rif 
cpots ill the an|!le« of the wslea. This very well diMcrilien 

ml sw thiw flowR near to Ilnlmna it iv very 
meriv<'r. (Cnvier & Valenrienmi- nay that 
ndiled AmUtipllln in outline, and pi«t«»«(il 
lome Miiecimena I have (ftg^ 12 and 13j. 


I». (V, II; A. H-, '.': -h-yrh ■^n-M.r't Ih-1.iw lii>t il-.n-al ^'■, i.'. 1..2.7; .le|.tli ..f .-amlal peilum-le 
:;'.2.ii]j li.'^i.l; s.-ali^L'TluU'!); i-.n-v 17 t..-.M lU t.i l:'.; In-h- hcuvv t..rwani. la]*ri[iK fnmi Ihe 
.-h..Lil.UTy tu Ih.- .■aii.ial ]»-.liiii.a,.; jaws li,-iivv. Ii|,^ (lii.'k; ^'lUM '."> t" 2.:> in Ii«i.l; .-yi- -t.r. t.>5 in 

..m-Ln.liviilii:il.): hidi.-j-t .(..r-al :in.l iiual rays n«.l.iiJi: l-as- of .■ati.lal; liiirh<'^l il"r>iil ray 4.-'. I.. 5-3 in 

tlie lonefli; Wt <lrir>iiil ray :i in tln' lonjic*'!; Iiili'rai iin<i cflii.lal yivitc ivini'pioii'ni:' in yi.unp, wliirli liavc 
» Hcrit-a <il \\jd\\ cnisfliar^: tun tii;lii Imrs ii^nully c-<inllnpiit rivtT tlii' lalfml i^\t>\; liriH ilii^iky. llie verli- 
cat iimv litilitcr-CKltn-il nnrl uiili Minn- i'|>fiti' on tlu'ir Ihimv ; no H|>iits on lii-a>l <ir iMHly in tlti' yrHin>!e»>t: 
in lar<!cr ••ni>:' i'|Hit!' n|i|H-9r nlHitit tlio (bim' ■>[ llip )M>ii<iral, ■ijM-nlc, ii[i-/k- <if ijnf'iH^n'k' an<t manilible- 
There in gniit variation in lli« ilit^inctni-wi •>( llic Inlrrul Iniiit)'. 



ItHh.lWM; tyiK- A hJlHlll<l^l■^)^^^■^wrlirtilll■t vi-niial Imni, 

I c'uiuIhI r']KiU Rn> <lL:'ttiirl. tlif ilark i-niwliiirs itn- iliirkfttit in 

trixls frcim tlii> I'vc lo Mi'iiriil <lark liar, iImh with tliu durkt^r 

<>rrii]i|c<l lutcnil ImiiuI: rlii-cks iiiiN|iiitl<il, njHTt-li'S iiml miimlililp 

; :i fi'H- Miilli-n-il hjkiIc hIhtii; llii- Hiikt;; viTlii'iil l\i\i' iiiiirn iir lesB 

iimMMlmntlH. ciii'li M-altMif 1li<- ^>ii1t' with a iliirk .iimt l'iiniiili^)i>ii);i- 

sly f|Krt(iHl; clii-ckM i[i tlu' younu "f l"itli ty|Kii iincpDtti'ii ; Hi'liy 
l«iti(il liiiiii ill tli.Miciiilt, Fip" 11 mi'' l-"'r'' -Inittn (r..m uml.T. 
VI' orvari.i in llu- «»iiif ctiiin' "f ilcvi'lcj[iiiH'nl. Fif;. 12, lyjio II, 
:i hy l*m>y. ami i.J Ihi-.iriirinal Mmmn(A«-. Fij;. 11, rcjiresi-iitiiiir 
w l.irrolhiyi F,ijj.'miiiinii. \ iiiiv. { No. itilTl', In.l. ITi.iv. Miii^. I. 
(78 B|lwam.w, 44 to IfK) inin. L.rij;. fr.m, Ssm Cristoluii. | 
M(«t of tlieM' wi>n- taki-i> i.iit nf » iiiii<l<ly l.i)r>H>ii iu»ir th<> nvi-r, nml all won' vi^ry t>al<' in n.lor, tli>< 
croKtiliarx Hlinwiii): wiill. In (Ik-h- jmliT r>|Ht'imi'iihi llicri' an' ii'i itiilii-iitiniis i>f ii liiii<;itiiiliniil t<tri)M'. 
Tln-brKOHt, wliirhnlHui-niiii-iiiitof llio 1h;;<><>ii, i» iitilrly miir<<riii liifht hhIiv. tlicn' lit'irie Iml fninl in<li- 
catiim "f I'roHMlMrH ami .'iioti'; IIuti' an' faiiil fi'iu nii lin^ ;tiiil ii|«ti'Ii-. OtluT Miiialti'r ti|>oi'iiiicn[< are 
every whpn- |>r()(ii!i(>ly H|M>tl<-l. In tlu' iLirki'i' i'|in-i[iifiiH tlu' rivcT tlipir i< a ilark laloml lianil. 

[ji till' larBprHiKi'iiiH'nr' tliiTi- an- 2 tj 
altcrnalin); li)rlil and ilark; tlic latcrul ii 
a liiii- iM'twivii tlic twill n lUrk iiin'ak c 

witli ilark i<|>i>tri; vciilrnl s 

H[HiIlnl at tlic Ikiw'. Tyi« 

tuilitial wric!'; chii-kp an wt-U a.-" ii]kti' 

fltrpAki<: wrtual linn imiro ciiiwiiiniijn 

of llid yiiiinj: nf ty|n' II liitw n-KiiIarly h 

of the Niiiii' wiitf atiil with rcproilufti 

oviili-ntly n'|>reM-iilH thi' vnrii'ty lijiiirc 

(711 -[H^l 

iiru'iiH, 17 t'> i: 

K) nil 

n. h.nii, fr 


"■'""'"■ ""yp 

.■=- A 

anil ltfn>i 


11) iiliini.' thi- 

tow:' of w-alcN 




ll Ihi-y PI iri'ail 





ihi^ larvi-ft fjHS-inii'df" ami r 

ii>l in 

olh<-rs. ' 


■i-loiMs! a- in 

wiim- Man An 


. H|Hvini.'n 

(lark HiMN-iincii n 

>HCinlikw in ah 

LLI>Ml i 

I.II r,.-.p-<-li 

■in.iiuiliml-, jicail 

(l~Rp(N:imens, <12 to I.MtLiin. Icmv-. from 

(4 BiHH^muiiH, Ji-'i ti> 240 mm. Iimt;, fniin ^ 

All tli(wc RjNM^inciiN an.- ilark, llio WMalliT oi 

and i-atiilnl H|Hit!' diHtiiu't, Tli<; larjicst mic if a in 

witli hlai'k Hlri'ukH and HjHjtx on rlx^^ki', 'i|K-n'lfM, 

Tliewc may rt'ferTeil Ui liiriiciiulhui. 


Aniiini: :V2 !i[«t'iiiifnf fniiu Siiu Ant'idio tlii'if art- '-i liiHtimi ly]Kii. i.iu- ni uhicli may sitnply U- Hie 
wlull 'if iiiit' .if ilu' '•tlii'r:'. Tlii'y iiri' ull f li)ni.iil*'. Hit- .li'|illi lii-iiitc 2.;> to 2.'> in Icntitli. There an-, in 
theliryt |iliui'. 4 iuliilt." iii'a-'iirLnjr 1'>lll<illN> tiiiti. in lin^h.cif tygx- IMmiii ('aliil«/^r. Tlify <liRer fniiii 
the CalalEizar s|><i'iiiii'iif> in ImviiiL' tln' Htxtls Hloni; tlii' naltt' lar^n'r ami kiw regukr. In twn nf the 
i'|HH;iiiU'iiM Ihi' I'hofkr' ar<' 8]><ilU-<l. In Ihh i.tlirry the s^kiIs nn- rnnMiii'iit Into vi'Hiiitl i>r loiitnliKlinal 
sticks. Ill Ihi-lnrwfitH.mriif lli.M|..r«ilniysiirv|,n.i..iia.-l. r.-.iiliiiij;1.. ntar ini<l.lli'..( mii.lal. Tiiew 
an- i.r..l« ih.' whilt ..f 17 si^rini.-Lis rr.iiu !I2 tn ir,7 luni. l..,iir. In all ..f Unw. .'Vf-n in th.. wuall<«l, 
Ihcrr nrc Njxiti' nn tho rliivki'. more niiniiTons in lh<- l:iri;i'r. Hti'l i-ontliinit into >>lniikri in Ilic tnrtnvt. 
Siili-s irninilarly; lal.-nil Umrl in..ri- .ir l.-ss .■..n^iiiriiniis ii." 111.- Ii.-<li is litlil.T r.r.lnrkiT^ there 
nn- tnn'i-^ 'if tlu' ii-nul lit-hl an>l <brk IkitkU in s<>ni>' i[i<livi.l>uil^^. All of lh<-»<- an' I'videnlly tvi.i<at 
//. /,/r.i.Y,„M.<-. 

Tli>Ti' an- II ;.(H-<-inii-ns, '.»\ U, IL':! iniii. .-vi<l.-nlly ni.Kiili.-ali.ins of //.'.inth'i' l»rr.illmM 
(n>iii<'alalHizur, In whirli, ex.fpt on.' vitv liiirk ^|>i^'ini<-ii. tl.f lal.-ra) kinii is vtry .-.inHpiciunis. Vcr- 
tii'al luinilM i|iiit(' wll nmrkul csivjit in ilarkpft s|>i'iiini'iis. Clui-ks ar.' nnsin.ttt-*! cxcei'l in .me 
inaivi.liial M-hi,'), has lainl ~|..I.s .-i.U-s »ith(>nt small <lark *>{K.ts. 

Heros tetrncautbus griseue Kiiicnnjann. iii-w snlis|>ii'iii:. ' 

Tv]..': No. INiTil. In.l. ri.iv. Mns. ; a -'ix't'inii-n 117 mm. l..nL-. fr Sun Am«ni<>. 

I»- irv. II: A, tv. >i; «:il,T= L'T; |>..n-« IT - 10- -li'i-lli ii lilll.- m.jn- lh:iii L'.5; .l.',.ll, .,f .■aiulal 
|><-<liinc'k' ->..'> in hcail; lit-a<l '.'.7 in Ilie k>n|:lh. Sha)K' aii'l p'iK'ntI .-liar.u'U-rs <>f y/,r'» h-lm'-'inlhw, 
from till' sum.- |ilai-.-. iliff.TJn^' in tin- cnl'iran-l the nntahly \:\tvkt i-y. ;tv <'nni]>nn'<l uitti >^)H.>.'imftu' of 

//.> ni Thi' sain.- si/<-. Ky<- :<.7 ill Ik'^.I ,4.'i in //.;„lh 1' thi- sanit- si»-i ; 1 in ililer- 

■irliilal |l..i); l.i". in sn.ait I 1.7 k l.r,-..rliital lin-si-Vfnlhs i.f i-yc: ^noiil ;l in iuii.h hiuln-sl .kirxal 
aii'l anal n-achin); lia-'i- nf I'anilal: liiiihiT't ilun^l 4.7 in ihi- It-nctli; hi|;hol anal 5; vi-lilrals n-a(-hin(> 
to vt-nl. No lateral ^i»t; a faini i':iuilal s|>ot; siilis ^L-hy villi irn-L'nlar >iark i<ii>il; a few uhltlxh 
stn-aki- [lir<in);h some of the smlii' alHivi- laleml line; i-h<-<-ks [ilain. a fuw 8\iota on ojierclefl; soft 
portions of vertical tins spotte"! nl !ia»-; no Irwer' of ilark er-iK-'ltiirv, 



Heroa tetracantbus latua Kie<.'niiii 
TyjK-: S.». iXKKI, rncl. Univ. Mu-' 
Thir' ix 11 narrow, <k-v]> llsh witli 

Itrolilc DvtT i^yt-". I), xv, U; A. iv, "I; 

porea, 18 i- 12; unont |n)iiile<l 2.7-"i i 

,n. n<.w ,„l»i,,.OT. • 

a Hini-iiiicii I'U) rrim. lonfc, from Shii Jiihii. 
nijci'litiK loH-iT juw, El {Miiiitol HiKiiit, anil a <lt'|>if>flHiciti in tlie 
l<-l>lli 1>; hi«<l 2.T; ileptli «f iwkUI i>o<liind(> t in iua<l; mhIuh L>!4; 
hiail; eye in hiMii, I.-") in iiitpn)rliitiil ; n<> diminen In 
np|>i.-r unifle iif fnll-i'pt'liinB; maxlliary n-ui'liinjf vdrtiral from front of orbit; hifihwit ilonwl and anal 
RiyH ntai^hint! to i-ml of liuRal lwo-liflliF> of ntnilal; hit;licHt clomal raj twin'aii \af)\ ns liu>t ray, (■qiialinK 
IrinKoft (niiilii) my, njnutinic K'nt^li of liiwl wilhoiil ci|H'n'li>; hi^ihu't anal my NliRlitlyHliortcr; vcntralti 

Asliy eray, ilarkcr nbovi' to liithl lii'low; tnii'Ii w^ali 
pxlj'nrlinft oviT In ncxl Brali' anil fiiriuiii); iliHiinct wrii* 
iif dioi-lt, anil lower jitw willi ilark fiimtc; vfrtiral linw 
ilrttk, iniiur my li>;lit; |iwtciralH lali-. 

Heros tetracanthus cinctus Ki|^■nnlann, new milwi^i^icH. ' 

Type: .No. m71, Ind. I'niv. MnP., ii s|>i»'inii-n 12<) nun. lone, fmin Thj 
Fonr Rp^n'inK'iiN w<'n' tak«n ill Vvsti Kcal iiiefliinriii): ii (>S, h 72, <■ I'J), 

lively, anil ■llftt^rinti in ruloration [rum IhiiM.' taken nt ;iiiy other point. '1 

one fiavint; very liislinct niarhiii^-s. 

' iif Iciwer [lartM of Hile with liliu'k Piiot on tip, 
; lower [wrt of ojH'ri'le, prifliK'n-le, luwer |iart 
ilnxky; rtift clorial anii aiinl n|iolteil; vuntralH 


1(1 if l:)li nun. 

long, ni>|iei-- 


n- very .lark, 

Thi' sipociuienc appmacli //. ui^iiiitm lint have a normal latenil line; ('(tlietyin-) Ih mmt al>ernuit 
ill ilii (■iiloratioii, li ajiproarlKW the roloratlon i>f IflTtuniMiu liirr-illniM from the Almi>nilurei<, x niiil li an- 
iniliKtinfniiijhalile from 'itlmr yniinj; cxi-0]>t in thi' Itanil throiitih the liilerni it|i>it. 

]). XV, II {r); xvi, 10 ('/); xv, 10 |/,); xvi, II; A. iv, H; n-aX>y* US, jHire.. 1.) + H {"); lll + S{/,); 
15-11* ('■); !*>-( 1' (''); li'«'' ^'i '" -■"; 'l''l>"'i -'-i'> t" 2.2; .leiilhiiteml of iiiHTi'le 2.J t.. 2.4H; i-ncmt 
liointiil, 2,75 to ;i in lieiul; month h'irizoiltal, niiixillary iieitrly eoneiitUil when the niiiiitli if eliineil; 
maxillary nlmnl n-ui'hini; vertical fniin fMnt of 'irhit-; hlnhitit ilursiLl nml anal rayii rejiehin^ little 
lieyomllw.ieofcmiilal.'iinlenKtli, ia.stiliir»al rayi:liii thelonKei'l; ventral;- muhint'vciil; ;iill^)|n-nintr 
witli a BUppienientiiry iBHicli alxivr, reneheil liy a larjn-r or smaller foramen. 



Ditrk; ii i.Urk liitiTuI :>ri<l a KUirk •■.iw\a\; yi.l.- uitli 7 liiflil .'ro^lnirs; a lielH Hinnk a.'r.iffi 

linlit; Hitic-i-U-twiTd liu;lil Ittrs ti.nii <l:irk Uirs ..f al-.iil cnuil w i.llli ,-\.t,>1 lin<l l»<>..ii U-iy, wliid. 
are jiuk-Ii »i.l<-r »l lli.- t.>,.: a stn-nk from fyi-aiTiw iipi^-r aiit:!.- i.r j.-il)-i>|H'i>int.' lo |U.> H-.-<>n<l 
ilnrk liar: i-ljc<-kH ami •>J>l■^l'll'^^ with liku'k sjhiI.'' ami sireiiks, ventral Hiirface sixilUil; viTlicat lilH Hiiil 
v<-tilratH'lark: soft tli>r^l ati>l vvtili olcx-nn' x|>ob>; iHit'inil yaic. \m^\Mitw<i. Tlii.- dark lian<l in 
wliii-li til.' latiral s[<,l in l«f»t<-<l .'niniiim-l l<> tli>- Uuk- 

YouiiK niilc-li liu'hl.T c'i,lnn-<l, ji few .lark »]«it.- iil.-iii; ..|nti1.. am! U'low cli.-.'k: v<-iitral i-iirfa.-.^ 
liri:'|K.1t>-'l. lalcnii ami rainlill si-it." •'i,ii:<|.i(-iiuiis. 

A Hftli .-]N-.-i.ii.>n, ir.:! mni. 1..J.-. fn,ii, Vx-^. l^.al. is a lypi.^al f./r...-.,.,//,».. 
HeroB ni^icsna Kim>ii[iiunii, m-u >:]Hi'if!i. 

T\t>< No •H><<.^ lixl \i\\\ Mii^ I i-'iiiiMi I'l-'tiiiii ]<>ii|. from liTiar. Ill Ri» 

Oi,. xiiiiiii.ii ttn-'iTiin'-t III lli< im.vl prriMiiiKnl >f th> iliprniil fonnr (mil -Ik.iiI'I iitiii.-i 
talln,.l^ .1.-^ nU il u- i .ll-lilir 1 "jk-. i.~ il 1 I> < I I>1 >irii-l iiii>r> tli <■> [■.-. line ri 

1) \n 11 \ i\ 10 -i-ii.". .'s |-,r.- il-iil 1> (I li«i.|-'(. I([.tli I infl. I.-*. lliiiN -' 111 tlif 
loiiklli I'l'tli It Mil) of ,.].ryU 2iiilint.t]] I liiiiKt.r .l<\, .\,^.^h \. r iiml.ll. of i tc iviimN 
IciikUi f li.'i<1 1.— om aiiiii(I<r -iioiit |..iiite'l iiftxr iii.l Inu, r | r..lil,-^ ■>. irk .-.iiiall^ 
imlimvl 1 IhIihiIim hi ]« r ].ro]il. ti>>>"»i U Inn 1 < i> , inixiM ir\ \>r\ Iil1l< < tiivxhI ulxn ith 


PhilypnuB dormitator ( I aiv pCnle ) . T. 

Rio San Jnan» im<l »een in Vonto nprinp* near Havana; v<»r>' abundant in thi» San Juan, where 
8I)e<!iinenH 4<)t<> 2rA\ nun. lonjr wert^ obtaintNl; younjr with l)lack strijn* from tip of lowor jaw to candaL 

Dormitator maculatus ( lUcxh ) . T. 

Rio San Juan, at it? nioutli and at tlie first ford; very <lark; lair<M)nat San Cristolial. v«»ry li^rlit. 

Eleotris pisonis ((rniHin). T. Rio San Juan, at its numth and at \hv iirst fonl. 

Lophogrobius cyprinoides (Palliis). M. Rio San Juan. 

Oobius soporator ('uv. Jk. Val. M. Mouth of Ric» San Juan at Matanzxi.^ and at its first fonl. 

GobiuB boleoBoma Jordan <V <Til))ort. M. Moutli of Ri(» San Juan and at it.^ first fonl. 

Awaous taiasica (Lichten.'^tcin). T. Suniiden). 



Stygricola dentatus (rcH>y). t* 

ThiH hlinil fish waH taken in the M rave noar AlacraiU's, 20 s|M'<"i!iu'ns; Jai«>ruan. T) siK^'inuMis; 
FriaH, 2 «|x*<*iniens; MinlcHta, 4 .-^ptH'i mens; and TnuKpiilidad, 7 s|H'rinu'ns; cavrs nrarCanas without 
HIKM'ific. hn'ality, 5 Hj^otrinienp. In all, 43 spocinirns were stM'unMl, ranjrinjr fn>ni <U) to 1.'>*J nun. lonjr. 
I'<Kiy n^'onls this HptHM(*8 fnun Cajio and Ashton, in whi<'h it was n<»t found hy us. Ih* al.«-o nn'ords it 
from I^ InduHtria, which is 8ai<l to \h}. l)etw<»en ( -ajio an«l th<' ('anas caves. 

The niak»8 of Stiftjirola dniOituA are distinctly larj^er than the feniaU»s. Avera^i* length <»f tin- 20 
females cjiu^ht is 97 j- nun., the larj^estone 120 nun. Theavenijresizeof the 21^ mail's is 11:5 inm., \\\v 
largest one l)ein>? 152 mm. lonjj. The males wen* in excess of the females in the nitio of KM) feinalrs 
to 115 males. There is })ut an appnn'iahle difference* in the averajjt^ of the fins, as far mi these <Nuild 
Ihj countinl, the averajjje formula for the femah« Ikmuj; I). 01.4, A. 74; and for the males I). 01.1, A. 7.1.0; 
or the avcra^ for the two are D. 91.2, A. 73.6. 

Lucifuga BubterraneuB Poey. t * 

This 8pecic8 was taken in all but one of the cavw in which Sty^irola (hnUihm waA taken, and in 
fleveral others Ixiflides. The localitiofl are Ashton, 13 specnmens; Los Baflos, 5 specimens; Cajio, 3 sin^ci- 
mens; Hawey, 16 sjiecimens; San Isidro, 2 spcn-imens; Jai^an, 18 spcH'imens; Frias, 5 siKHMuiens; 
Modcsta, 2 specimens; Tranquilida<l, 3 specimens; Caflas, without sp4»cilic lcK»lity, 9 s|HH'imens; toUil, 
76 spt^cimens, ran^U); from 24 to 94 mm. long. The femali's of Lucifugn ttuNnranntH an* distinctly 
larj^T than the males. In making the average for the size of the sexes individuals less than a year old 
were not considennl, lKH*ause the differences in the sexi^s wouhl, if present, Ik* hut very slight, and 
bcH-aiise in such young the sex could not always Ik» determine*! with a»rtainty. An exannnation of all 
si>ec*imens makes it prolwhle that at the end of a y<«r aftt*r birth the young an; alxmt 50 mm. long. 
In obtaining the average size of thc^ st»xes only s[)ecimens over 50 mm. wen* considenMl. The maU*s 
al>ove this size measun* 59.7 mm. on an av(*rag(*, with a maximimi of 94 mm. The femaU*s mt^asiinMl 
71.1 nnn. on an average*, with a maximum of 93 mm. Of the sptH*imens over 50 mm. long 2:i wen* mah*s 
and 22 femah»s, or 100 females for everv 104.5 maU^. The fin formula to the mnirest decimal for those* 
of the individuals over 50 mm. which wouM Ik* counti*<l is, males, 1). 82.1, A. 67.4; femaU^, I). 81.9, 
A. ()8. Theaverag** fonnula for those l(*ss than 50 nun. l«mg is 1). 83, A. 67.2, or for all together, I>. 
82.6 f, A. 67.5. 

While the? average nundKTof rays differs considerably in the two 8iK*<*ies, thenumlK*r in eae*h varies 
so much that the numlK*rs in individual cases overlap, individuals of Lucifutja n*a(*hing as high as KS 
dorsal rays and imlividuals of Stfftjirtfln as low as 87. The same is true n»ganling the anal. 

A female of this 8pe*cies (fig. 3, pi. 21), 65 nnn. long, containi*<l four young alnnit 20 mm. long. 




Contributions from the Biological Laboratory of the U. S. Fish Commission 

at WckhIs Hole, Massachusetts. 


l*roft 'S ^t tr t »/" ZtXiit i^y in I h'N ison f '» ivt 'rsity. 



iiitr(Mlu('ti(iii -3i» 

Rt^vitiw of lit^'nitun^ 241 

Teriiiiiml buds aiul tlieir iiiiHTvalinn 247 

iMinctioiis of UTiniiial IhhIh 250 

KiXiK*riiiu*nta on wiluroid lishi^r* 25(» 

KxiKTiiiK'nts on ^^cidnid li.**ht*s 257 

Tlie liakis f 'rophifns tnntis 258 

Tlie t(>lll<'<Ml, MirrinjniJits timinul 2()2 

( )tlH?r fislu's 2«>4 

The w.*a-n)])in, l*rioiuttuK iiintHnna *^\A 

The kinn-lisli, MtntlrirrhuH snxfttilis :i(>5 

Till* toad-l'lHlK Ofmiiiiin tnit 2<»5 

(;oiif:lu»<iun :^(k» 

Addendum 270 

Lit4Tatiin> <ited 272 


Contributions from the Hioloi^ical Lalniratorv of the U. S. Fish Commission, 

Wmxls Hole, Massachusetts. 


liy C. jrnSON IIERRICK. 
Pnt/tssitr of Zoology in Drrtisofi University 


The pnu'tical problems t-oiinected with the tisheries have been atta<*ked (and in 
lar^e iiiejisure siu'ces.sfully solved) hv a roiij^h -and -ready application of the method 
of trial and (»rror, and th(» >eientitie investigator has merely to follow after and 
exphiin why a given form of tnip or method of lure is successful with one species of 
tish and not with another. But there remain many unsolved problems of great 
economic imjx>rtance, and it is the function of scientific research to contribute to the 
solution of these problems in a more orderly and economical manner, even though it 
often happens that the investigjitor l>est (jualitied to solve the scientific problem has 
not tlie practical knowledge of fishery matters necessary to apply his own results to 
economic problems, and so his facts have to be worked over from the other iK)int of 
view before they become practically useful. 

We are, in fact, profoundly ignonmt of the senses and instincts of the fishes, 
even those connected with their feeding habits, which are of so direct imjKjrtance to 
all commercial fisheries. Nearlv all which one finds in the scientific literature Ix^ar- 
ing on the senses of fishes is merely inference of function based on a study of the 
structure of the organs — a most precarious pathway for scientific research. My 
own studies on the nerve components of fishes have led me to certain inferences 
regarding the functions and the distribution of the orgjius of tiiste in fishes, and- the 
present study is an attempt to follow out these inferences by the determination of 
more exact facts regarding the pathways of gustatory stinmli a.s anatomically demon- 
strable, together with sufficient direct physiological exj)eriment to furnish definite 
information of the function served by this system of sense organs and of their 
nervous paths in the fishes. 

Neur(»logists have always jiaid a gieat deal of attention to the conduction paths 
within the central nervous system, and in recent yeiirs sj)ecial efiorts have U'en made 
to isolate the various functional systems of neurones, tracing the exact path of the 
sensory impulses from the peripheral organ to the primary sensor}' center, thence to 
the various secondary centers and return reflex ])aths. This motive underlies tlie 
recent studies on the nerve comjKUients and, indeed, much of the l>cst morphological 
work on the nervous system in all times. 


Some years a<^o I f(»rinuhite(l tlie followinj^ definition of sucli a funetional system 
of neurones, with sp(»cial reference to tlie iwriplieral members of tin*, sN'stem: 

Tlie Hum of all tlie iu*rw liln'rs in the IkmIv which imm^hosh certain physi(»l<»j;iwil ami iiHirphol<>)i;ical 
characrterH in common ho tliat they mjjy react in a common m»)<ie. Morj)holoj;ically eu<'h Hvy^tem is 
defineil by the terminal relations of \tn lilHjrs, l)y tlie orjijanH to wliich they an* n*late<l iieripheraliy, 
aii<l by the centers in wliich the liln'rs ariw or terminate. The fillers of a hin^le nyntem may apj)ear 
in a larjre mmil)er of nerven rei>eate<l more or lens uniformly in a metameric way (as in the j^Mieral 
C'utaneouH wystem of the spinal nerviv), or they may all )>e conctintrate<l into a sinj^le nerve (a.^ in the 
optic nerve)- 

Now, if we add to tliis the secondary paths relat(»d to the* i)rimarv centnU (»nd 
stations referred to abov<% and the chief reflex arcs din^ctly associat(»d ther(?with, we 
shall have a pictim* of the system in its entirety. 

Tlie functional system with which we are esj>ecially concerm»d in the i)resent 
research is that known to comj>arative anatomy as the communis system, including 
(1) unspecialized visceral sensory filwrs ending free in the imu'ous surfaces of various 
visceni without special sense organs probably phylogenetically the more i)rimitive 
elements — and (2) specialized sensory libers always ending in conn(H*tion with highly 
dilTerentiated sense organs in the mouth, pharynx, lips, or outt^r skin, known as taste 
buds, termiiml buds, or end buds, and in geneml serving the function of taste. These 
specialized elements are probabh' of more recent phylogent^tic origin than the lii*st 
group, and the term ** gustatory system" will be used to designate these organs, wher- 
ev(»r plm-ed on th(» body surface, togeth(»r with their nervous pathwa3's toward and 
within the brain. In other words, the gustatory system is that portion of the com- 
munis system of neuronics which serves the s(?nse of tjiste, as distinguished from 
those communis muirones which serve less highly specialized visceaiil sensations. 

These two grouj)s of fi!)ers can easily l)e distinguishe<l peripheniUy of the bi-ain, 
but centrally they have not as yet been suc<'essfully analyzed. Hence in treating of 
the centnil gustatory ])ath we can not be sure that we do not include the unspecialized 
visceral system also. But since in some lishes the gustatory fibers i>re[)ondenite 
many fold over the unsjH^cialized libers of the comnnniis system, there is no 
ambiguity arising from this central confusion of the two elements so far as the 
gustatory system is concerned, since the secondarv |Miths as clenrly tnu^eable in these 
fishes must be made up chiefly of gustjitorv libers. 

The central gusbitorv path is not definitely known either in man or in any other 
vertebrate, so far as shown by the available litemture. I Imve tlu'refore studied with 
some care the brains of some lishes in which this syst*'m is enormously developed, in 
the hope that they would throw light on this unsolved problem of vertebrate anatomy. 
And in this I have not been disappointed, though my study of the central jniths is not 
yet sutliciently advanced for i)ublication. 

As intimated alK)ve, sense organs Indonging to the communis system and pre- 
sumably serving the function of tiiste are found in the mouths of all lishes ('' taste 
buds"). They are frequently found uiK>n the lips, and ii. some csises they are found 
likewise plentifully distributed over extensive areas of tlu^ outer skin of the head 
and trunk. In this latter they are commonly termed terminal buds or end buds 
{/UiMfifKsptfi^ Ii(vhrronjant\ of the (iermans). They must in all cases be shari)ly 
distinguished from the neuromasts, or organs of the latend-line system ((lerman, 

THK <»K<;AN and SKNSE of TASTt: IN FISHES. 241 

Xt'rr*nhi}ff»f)^ tbouj^li tlii»se liittor occur in tho skin of tishes in a ^reat variety of 
forms, often rcscnihlinjr the terminal buds verv rloselv. The innervation and 
fuiu-tions of tlie two svstems of or<ran> are, however, whollv different, and tliev reallv 
have iiothiiiiT to do with each other. 1 shall illustrate more fullv in a later siM-tion 
of this jKi|H'r the structure of the terminal huds and the details of their innervation. 
I here call attention merely to the im|M)rtant fa«t that lH)th in structure and in nerve 
supply they resemble most closely the taste l»ud^ of the mouth. From thi^ one 
naturallv infers for them a jrustatorv function. Since, however, inferences are not 
in order when facts are available. 1 have undertaken to determine ex|H'rimentally 
the function of these cutaneou-; sense oiirans of the communis .^vstem. 

The exjH*rimeiit> which 1 have made are of an cxceedin;^ly simple nature, the 
attempt beinj^ to put the tish while under ob>crvation in as nearly normal conditions 
jts j)ossible and to utilize the ordinary feedinjr and other instinctive reactions so far 
as jMissible in the accuundation of the data. These are the methods of the old-time 
observational natural history, it i> true, as contrasted with the methods of pn»cision 
of the modi»rn physioIo<^ical IalK)nitory. They have, however, proved sufficient for 
their purjKjse, which wits merely to determine the class of stinndi to which the 
terminal buds are s(Misitive, or the sensjitional ino<lalitv which thev sen'e, nither 
than t<» contribute* to the chemical physiology of tasto in crenenil. 

The chief obstacle to exjx»riments of this sort, and one which many observers 
seem to have made no serious efforts to overcome, is the natural timiditv or shvness 
of wild creatures when kept in the confined and unnatural (piarters necessary for observation. The role nlaved bv fear in animal behavior has l>een vividlv 
brou^^'lit to our notice by Whitman ('1M>), and. like this observer. 1 tind that youn^ 
animals which have been reared in captivity are much more approachable and 
tractable under ex|>erimental condition.^ than adults which have lK*en reared in their 
natural freed<Mn. In fact, with s«'veral species 1 quit(* failed to <^et the adults to 
take fo(Ki at all in captivity, thouji^h they were under obM*rvation for lon«^ |)eriods. 


Surprisinjrly little attention has iK'cn paid to the physiol<»t^y of tjistc in fishes, 
and this literature i> v«'rv s<antv. On the other hand, the anatomical i n vest i<rat ion 
of the^^e sense or<rans has ln'cn ext<'n>ivelv followed for iiearlv a centurv, thiiuirh 
often in a blind and profitless way. The history of opinion \\\nn\ the sij'-nificance of 
thes4' sense orjoin> ha^ U'en ipiite fully iriven by Merkel C>>n) in his ^reat mono- 
jfrapli published in Isso, and the earlier plia>e> of thi> history need not be ajj^tin 
reviewed further than to mention a few salii'iit feature^. 

In X^'ll \Vel>i'r observed the ttiste buds on the |MMuliar palatal orj^an of the carp 
and correctly interpreted their functi(»n. He also fi*rured the brain of the carp, 
illustratin<4' the enormous vajral lol>es from which these ta^te* buds receive their inner- 
vation. I^'vdii^ discovered in ls"»l the terminal buds of the outer skin of fishes and 
j^ave a <U'tailed ac<'ount of their structure, which subscMpient research has shown to 
Ik» in some res|M'cts inaccunit*'. In \Sku\ V. E. Sihulze j^ave a more accurate descrip- 
tion of the *"^nrhtrfnrmi4jt n Onjiin* " of fishes, in which he distinj^uished the s|>ecitic 
sensory cells from the supiKji'tin*,*^ cells. He also correctly inferred their function to 

F. c. B. lyu^-l6 



Tubercnium acusticum^' 

>. Facial lobe 
f^^lobus vagi 

be sitiiilar to that of taste buds within the mouth, viz, the iK»iveptioii of eheiiiical 

In 1H70 the same author (F. E. Schulze, 7n) made a further im|x>rtaiit rontri- 
bution to the problem of the terminal buds by the demonstration that thc\' differ 
structumlly from all neuromasts, or or^ns of the lat(»ral-liiie system. The neuro- 
masts are commonly sunken below the skin in eanals tulws, or pits, but in some 
eases thej' are strictly superficial and res(»mT)le in external form tlie t<'rminal buds 
very closely- a feature which led Leydi*( fol, 7t», 'JH) and otliers to assume that the 
two classes of organs are mere varieties of a common tyiM». Schuize showed that 
the neuromasts can in all castas 1h^ differentiat^'d from the terminal !)uds l)v th<» fact 
that their s|HH'irtc sensory cells (|H»ar cells) exti^nd only ]wirt way throujrh the 
sensory epithelium and fail to reach the internal limitin<r memlirane, while in the 

t<^rminal buds ImHIi s|MH'itic sensory 
cells and sup|M>rtin(>f cells |miss through 
from external to internal limitintf mem- 

This distinctii»n was ccmtirmed by 
Merkel Cs<>), who, with curious incon- 
sisten<*v, wliile recoj(nizin<j^ th(» struc- 
(uml dissimilarity of the twi» classes of 
orj^ns, nevertheless, as we shall see 
Udow, ascrilx^s to lK)th essentially the 
siune function, tom'h. This matter 
was put to the decisive test in my 
contribution on Aiin'!urui< ('<>!), a ty|H» 
which iK>ssess<\s ]H>th terminal buds 
and neuromiists in j^reat abundance 
and diversity of forms. Schulze's 
contention is supiK)rted lM)th by the 
structure of the or«;ans and by their 
iiuiervation, f(»r 1 have shown that 
all neuronntsts i»f whatever form are 
innervated ))v acustico-lateralis nerves from the tulN*rculum acusti<'um of the !)niiii, 
while all terminal l»uds, whether within the mouth or in tln» out4»r skin, are imier- 
vated by comnmnis nerves related centnilly to a sinjifle center within the hniin. 
This centi^r is biloljed, the hdms vagi receivin<f most of the comnmnis tibers froui 
the mouth cavity by way of the vagus and glossopharyngeus and the lobjis faciali> 
the communis tibers from the terminal buds of the outer skin l»y way of the facial 
nerve (cf. fig. 1). 

Similar terminal buds have lieen found in the outer skin of many spi'cies <d' 
Telcostomes and in Cyclostom(\s, but, so far as certainly known, nowhere else among 
vert<>bnitcs (save on the lips of some other classes). Th(»ir distribution among the 
fishes is very irregular, l)eing most abundant among the siluroids, cyi)rinoids, 
ganoids, and cyclastomes, in genend liottom fishes of sluggish habit, often living 
in nmd and rarely l>elonging to the predaceous types which find their food chiefly 
by the sense of sight. The following list of fishes which have been shown to iK>ssess 

Fio. L— Dfirnal view of i\\v bmiu of ilic yell<»\v cul-lisli (hi^- 
tnim tflivurin Rnf.). Tbc olfm-tory Imlbfj witli most i>f tlirir 
enini hav*? In*i-ii n'luovwl, hIso tlio iiu'inbniiMMiM nH>f of tin* 
fourth ventricle. exiMwiiif; the fai'hil un<l vhkiiI h)tM>>. Thi.>< 
vciitrU'le is IxMiiKlod behind by a tninsviTxe tU\kv contain- 
ing the eomniiNtnra iiilimu Ualleri and the e(»nnnissural 
niirleiut uf Cajal.,'.'J. 

THE okc;an axd sense of taste in fisues. 


tonninal buds on the outer skin is hy uo means eoniplete, hut will serve to illustrate 
the wide i*aii<^e of species whieli have ac<juired this j)eculiarity: 

Arrriua. On tiiiis and IhhIv (Merkol, '8()). 
ArifHUiuT stnrio^ j*tnr>nM»n. Hn Uarln'I ( Mcrkt'l, 

*«()). AIh> other j*tiny«M>n.<. 
Atfonna rainphiartim^ jx^V^e. Hn \\\v villiforin 

tentacles iRMicath tlie lu-a<l ( liiiteson, ''.H»i. 
Ameiuriis imlns^ cat-finh, and otiier North Ameri- 
can Sih»ri<iji-. On barMt*tf< ami n«*arly iht* 

whok* IhmIv surfatf (Ilerrick, '01 K 
Amin ntini, Ixiwtin. On j-kin of head and othi-r 

I»arti^ (Allii--, M»7). 
Ainfuilln rn/f/tiris, (td. Hn th»* fins, lijus, and antt- 

' rior n(.«tril ( Merkel, 'SU; l5at<»son, *;H)). 
A,<ffiuj< a/hnniiM ( Mc*rk«'l, 's<j). 
BnrhuA fliirinfifis. ( )n lKiH>l«'t ( F. K. Sdinlzo, '♦>.'» i. 
Bniiichnnitnmn InmyitftUnm =r: ^ | mphinj its Imnu ttinfns, 

lanctdft. On th<*<»nil cirri (Mi^rktd, 'si)). 
Cnrnmi'ni:i nnrnlnn^ ;;nld-tish. On tlic whoK-lni^ly 

( nnint'roib* anthors; Ilerrick ). 
Vt^fthninontthiis = Ojhitij* ^VaW/i^j tlvin;: irnrnard 

(Merkel, '80). 
i'ltttux ^rrrpittii^ t^eulpin. On tins (Merkel, 'SO). 
i'ljinmr'npn = Otrriwi (Merkel, *80). 
f'tfpnmis (ytrpiOf ^^iirj*, and other cyprinoidi?. On 

whole iKKly (Merkel, *80, ami others). 
Ikuiifh}4*nrtu* (Merkel, *S0). 
/)M<v^/m///i»'x //(/////f, Indian <'arp. Over !he wlmle 

Ixxly fiurfai'C (Ix.*ydig, *{H). 
Eiirftehfo/tux = Mott'l/Hy U}nT-hk^Tt\oi\ rcK'klin^'. Hn 

)>arhlet«and indvic tins ( Hateson, 'IK)). 
(iatfnx ralltirins, co<l. On lij»f>, l>arl>el, fins, and 

iMHly (Merkel, 'SO; Ilerrick, '(K)). 

f.'tnln.^ lust'u^^ \umi\nfi. On the lijis, liarhlet, and 

judvic fins (liateson, 'W). 
finihix intrlniKjuit^ whitinjj. < )n lips ( liateson, 'IK)). 
(,'fniiii< iHtllar}iiii.t^ iM»llack. On li|>s ( l>iiteson, 'tK)). 
ff'aitlrnj,jairii)< = Mittilln^ thrive-! H*arde<l r(K'klin);. 

On all the harhlets and jjelvic tins (Zincone, 

'7S; lUitenm. '90). 
(iiih'ms^ p»hy. On fins ( Merk«d, 'SO). 
/fiplt'Htiiiipiis, sea horse (Merkel, 's<)i. 
LrptiH-rjthnin.'i ri/u'f*'i\, c<>n;:cr eel. ( )n ihe outer 

ami inner li|»s (ll;it«-s<»n, '!«>). 
Lt uniJiitlns ihiintiituj<. On the ImmIv irenerallv 

(U-ydi^r, '94). 

1^ ut'isriin fl^fhiihi (Leydi^, '•")7). 

//>//! riilijm'in^ \\i\\i. On harhlet (Merkel, 'S<»). 

Miilhi.^ ImrUitiis^ mullet. ( >n barblel ( Zincone, '78; 
Merkel, 'S<>). 

I'tlnmuizon JinnntUijt^ lamprey. On skin of whole 
Ixxly (Merkel, 'SO, and others). 

piitjitstt iix = (iiititinjiUi'U'i imuijitiiiity stickleliaok 
(Merkel, 80). 

li/tinhns (imanm. On the Imj«1v j^enerallv (Levdig, 

Stnrp:viia (Merkel, '80). 

Sllunu* ffifinix^ cat-fish (Merkel, '80). 

S^jlrn villi fans, sole. ''Contrary to the natural 
presumption, tlie villi on the lower (left) pide 
of the hea<l <lo tmt liear sense organs, though, 
iis Mr. Cunningham informs me, sucli organs 
are fonntl l)t*twei*n the villi " (Bateson, 'iK)). 

Tmrn rndfarl.*, tench. On l>arhlet (Merkel, '80). 

As already sujr^sted, our knowled*re of the funetions of all of the sense organs 
of fishes is very iin[M'rfect, since* s|)«»eulation based uiK>n structure has seemed more 
attnietive to most authors than accurate physioloj^ical research. The monogniph of 
Merkel ('s<>), with its ^reat wealth of accurat<* aruitomical data on the structure and 
distrihution of terminal buds in all classes of vi'rtehrates, o-ives an excellent illustm- 
ti<in of the dan«rers in the {with of even so skillful an observer when he goes Ix'yond 
the IhiuiuIs of observe<l fa-t and enters the field of sjK'cidation. This author recog- 
nizes the ch)se structural resem)>lance In'tween these organs and the luidoubted organs 
of taste in the human Inxly. He controverts, however, the clear argmiient of F. E. 
S<*hulze for their gustiitory function on merely theoreti<"al grounds. His first objec- 
tion is Iwi^ed on their iimervation. Instead of being supplied by a single gustatory 
nerve, the glossopharvngeus, they may be supplied, he says, by any c»ther IkkIv 
nerve. This objection has Ik^cii totiilly removed by the dis<*overv (compare es|M»cially 
my own Amtlnritx |KHH»r, already referred to, publish(»d in Octolier, VM)\) that all 
terminal !)uds, no matter where located on the ImkIv and no matt<'r from what nerve 
bninches their innervation seems to come, are in reality supplied by nerves of a single 
physiological system, terminating in the biiiin in a single center — the communis 

Again, he objects to Schulze's theory that the tenninal buds serve to XiMf^fifl^. 
gustatory stimuli on the various jmrts of the IxkIv, on th(» ground that an orgitn of 
chemical sense stimulated by substances in solution in the environing fluid eould not 


roceive a .siifBcieiitly circunisiTilied stimulation. It is uiiiiec<\ssary to follow tlie 
argument in detail, for the experiments which I shall describe shortly show conclu- 
sively that when the sapid suhsUmce is l)rou<^ht into contact with these organs or 
very near t*> them the stimulus is accurately and very promptly lo<'alized,and in fact 
some of the fishes studied ha))itually find th(»ir food by this very ])ower, the gusta- 
tcny stimulus calling forth an immediate reflex movement toward the point stinm- 
lated. It is probable that the loc4d sign is not given by the gustatory (communis) 
nerves, but b}' the accompanying tiictile (genenil cutaneous) nerves of the corre- 
sjx)nding cutaneous area (which general cutiineous nerves Merkel, curiously enough, 
denies to the lishes altogether, whereas, in fiu-t, th(»y are plentifully supplied to all 
|)arts of the skin), though my experiments do not decisivi'ly answer this (piestion." 
Weak stinmli, esiM'cially when unifonuly dillused through the water, are, it is true, 
not at all localized; but strong stinudi are un<iuestional)ly localized by one method 
or another. 

In fact, Merkel agrees with JolxM't that the terminal buds of the outer skin are 
tactile in function. This is based largeU' on the erroneous belief, referred t>o abov(», 
that there are no free tactile nerve endings in the skin of tishes, and also on the 
observed tactile sensibility of the barblets and other parts of the body known to be 
most plentifully supplied with terminal buds. Hut I have shown that all of these 
parts of the body receive, in addition to connnunis nerves for the specialized s(»nse 
organs, a most liberal general cutant^ous innervation foi tactile sensibility; and the 
experiments which follow go to show practically that these two functions conunonly 
coopc»mte in setting off the reflex of seizing food, though they may be exi>eriment- 
allv isolated. 

Merkel now proceeds to carry his argument to its logical conclusion (and like- 
wise to a rttdtietlo (id ahmirduin) by denying the gustatory function to all terminal 
buds, even those within the mouth supplied by the glossopharyngeal nerve, of all 
vert^^^brates l)elow the MahtntaJla, 

He finally concludes that both the neurojnasts of the lateral-line system and the 
t<?rminal buds are tactile orgjins, the buds !)eing the more delicate; Init if these arr 
deficient, then the neuromasts may be elevated to a more <lelicate functional value; 
l)oth of which conclusions, in the light of our present knowledge, illustrate the dangers 
attending an attem[)t to determine function on the basis solely of observed structure, 
without adecjuate physiological control. 

The general works contain numerous references to the subject, but usuallv 
chance observations or siH'culative conclusions. Ciiinther says, under the caption 
''Organ of taste'': 

S<»iiH^ lislics, esiK-^cially vejretalile fiHMU'i>«, or thosi' prnvidnl with broiui inolur-likr troth, iniu^ticate 
their ffMxl; an(i it mtiy Ix' oIxhitvimI in carps uikI other cvpriiinijl lish tliat this pnMi'.ss <>f inasticatitui 
fnMjut'iitly takt« some time. But tlie majority of lish swallow their f<MMi rapitiiy ami without mastica- 
tion, and tlierefon? we may conclude that tlie sense of tiiste can not l)e aeut<'. The tonj^ue is <»ften 
entirely absent, and even when it exists in it« most distinct state it c»>nsists merely nf ii;>:ament<Mis or 
cellular sulwtance, an<i is never furnishiHl with muscles capa]>le of j>nMiu<'in;r the movements of exten- 
sion or retniction, as in most hij^her vertebrates. A j>eculiar (»r;ran on the rouf of the palate of cypri- 
noids 18 iH'rhai)« an orjjan adai)t4.Ml for jjerception of this sense; in these lishes the palate between and 
Mow the upi>er pharyngeal lK>nes is cushione<l with a thick, soft, contractile substance, richly supplied 
with nerves from the NiTvi vakils and jjlossopharynjjeus. 

«0n thl* ]K)int, boc the further experiments reeonled iu the AUileiiduiu. in>. ■J7l>-.r71. 


Ro<^r(linjjf tho peculiar i)alatal or*^n of the eyprinoids, it has been known since 
Wel>er's account in Is^T that this is phMitifuHy supplii»d with taste buds, and \Vel)er 
hinis<^If brou<rht forward strontf indirect evidence that its function is ^ustator}'. 
The followinjr observations (and many .similar ones mijjfht l>e cited from the litemture 
of s|)ort) are taken from the section on ''TheTrouts of America," by William C. 
Harris, in the American Sj^ortsman's Lil)i*arv. 

Tlie anirlcr can not n*sist th<* U'lit-f that thr seiis4's t»f snii'U an«l ta-«»t4* an* weli tU*velope*l in trout. 
Th«*y «*j«*<-t the artificial Hy, if the h<M.k is not fa>t in the tlesh, at th«» instant tliev note it8 nonetlihle 
natnrc, or w hen they feel the jrritty impact of the hiM»k. They will not eat inipnre f<HMl, an<l they 
liave the faculty of iK'n-eivinir «M|ors, an«l varions s<*ents attract or reiK'l them. This has Ihhmi veritieil 
from the ♦iirlitvt <!ays of onr art, when ancient nuismen nsetl diverse* an<l <'nriou.« jwL^tes an<i oils, which 
wen' s<Hinctiv<' to tish; in Walttm's <lay, ami lonjr after, this pnu»ti«-e wa.« foll«)we<i an<i tlie re<-^)nls tell 
ns of it,*^ sa<"cess. When I wa< a l»oy an«l the S<"hnlkill Kiver wafj swarming with tlie small white-l>ellieil 
csit-tish, than which no more <ieli<;htfnl hn^akfa.**! foo<l ever cjime out of the water, the only Itait iii«<l 
to catch them wil** made of IJmhnrjrer <-lHH*f5e, mixe<l with a i>at<*h of cotton l»attin^ to hoM it firm c»n 
the h<H»k. No other Inn* hati the same attra<'tion for them l»e<rimst», no ilouht, of the ilecide<l cnlor of 
the clie<»*«e. 

The problems connected with the relative sicrniticance of the several sense organs 
of the fishes have !>een treated !)oth anatomically and ex|XM'i men tally in the excellent 
|>ji|x*r of Bateson ('i*o). After anatomical remarks,<l laro^ely on his own careful 
studies, on the eves, ol factor v or<nm>>. Jind ifU'^tatorv or<nins, he recounts a .<^ries of 
adminible and well-<'onsidenHl ex|>eriments ma<le to test the parts played by these 
orjnins in the normal feedinj^ of various kinds of fishes. 

These ol>servat ions are jjcrouiMnl under two chief h(»ads, viz, "^Senses of fishes 
which seek their food bv sccMit" and "The senses of fishes which s(»ek their food 
1)V sitrht." Thoucrh the taste Imds in the mouth and outer skin are descril)ed and 
correctly interprettnl in the anatomical \y,\vt of the paper, these or<rjins are scarcely 
considered at all in the physioloj^ical part, and this i> really the jjcreatest weakness of 
the pa|)er. Since my own ol)servatinns in j)art follow so closely in the footsteps of 
l^ite^on (thoujrii compl<'t4»d in the main iM^fore iiis pa|XM* was accessible to me), and 
.<ince they are iti <rcnenil confirmatory of hi>, it will l>e of interest to review portions 
of his paper at this time. 

II«* ofiv(»s the followiuiT li*^t of tislh»s which h(» has observed ''to show conscious- 
ness of f<MKl whi<-h was unseen bv tlxMU, as, as will hereafter be shown, there is 
evidence that they habitually >eek it without the help of their eyes": 

I*n4nitleni» iiuinrhii.^^ mu<l-tish. 

Si'ifUinni riimirtiln, ron^h <loj;-tish. 

St'iiHinin rnltilim^ nnrst^honn<l. 

llnjn fmtl.^^ skate. 

(\nnfrr niltjtirl.'t, confer e<'l. 

Aiiiptilhl mlijiii'is, mA. 

Mi4tlhi tnrirrntn^ thn^'-ln'-anleil nM*kIin^. 

}ffft*'!ia intiittrln^ livc^beanle<l n»rklinjr. 
Xniitu'hrUuM tmrlmtnhi^ liMldi. 
/ IjefnultHjaMtr (jnuanii^ sucker. 
Sthti rtiit/ttriitf sole. 
St^4'4i inimitn^ little p<^)le. 
ArifM'itstr nithciiun^ sterlet. 

H<» savs: *'To this-list mav almost certainlv 1m» added the remainder of the Haiidit^ 

• • • 

together with the angel-fish (RhJna sfpnifitm) and Tnrjpcthf,^^ Unfortunately, how- 
ever, lljiteson in his list does not distinguish l)etween those fishes in which smell 
obviously plays the leading |vart and those in which taste or touch or lioth are used to 
com|)ensjite for the reihiction of vision, and it is this defect which it is hoped that the 
present contribution may in part correct. 


Most of tlio forms in tho list alK>vo are more or loss nocturnal animals, Init thej'^ 
differ much in this ro^iinl. Tho part attributed to the sense of si<^ht and smell in 
Bateson''s studies is so similar to mv own conclusions in many respects that it seems 
fittint^ to (juote tho orroater part of his description, especially since tho sptH'ies 
observed by us are in all cases diffon^nt. He savs: 

None of thost' fishes over start in (|m^t of f(MMl wlien it is tirst pnt into the tank, Imt wait for an 
interval, donbtlens until the H4*(»nt has lKK»n <liffus«Ml tliron^li the water. Having iH'n^»ive<l the went 
of f<MMl, th(»v swim vajrnely alnnit an«l a])i>ear to seek it hy exaniininjr the wh<ih» area iK*rva«le<| I)y the 
s<"ent, liavin^ seeniinjrly no simis«» <>f the «linM*tion when<H> it pnH'c^tMls. Thonjjh sonu* of th«*H<' animals 
liave iindonhtediy s<»in<' visual jK>re(»ption of ohj*H*ts moving in tlie water, yet at no time wju* tliere the 
sli^lit4'Ht indication of any nM'ojrnition of any f<KMl suhstantH' hy sijriit. Tiie i»nH'<»ss of sinireli is i*<|ually 
in«linH*t and U»ntutive hy day an«I ])y ni^ht, whether the f«M)d is exiK)se<l or hidden in an o|>iU|ue vesnel, 
wliHlier a pi<M'e of aetual fo-^l is in the water or t!ie ymn*. only, s<ju<»ez<»d through a rh»th, and, histlvt 
whether (jis tt^Unl in the <'Jis(» of tluM^mtjer anjl the nK'klinjr) the fish In? blind or not. * * * The 
|H're«*ptions, then, hy which tlu*s<* animals recojrnize the i)resen<*e of fjwMl an» eh*arly ohtaimnl hy 
mftins <»f the olfaittory (►r^ans and apj>arently exclusively thnni^h them. I was parti<'ularly surpriseti 
to find n«> indication of tlie jK)ss<»ssion of su<*h a funi'tion ])y the 8<»ns«» orjrans of the Iwirlh'ls and li})H or 
hy those of the latend line. As has lH>en already <U»scri])ed, the iH*lvi(^ tins and luirl)els of the nn-k- 
linjrs ( hfitti'lln) and the lips, et«'., of mj)st fishes l)ear j^»at nnmlK»rs of sens<» orjraus closely com|tiinihle 
in stnn'tur(i with the taste- hiids of other vertebrates. No one who hafl s<K»n thi' mfxle of ft»e<linj; of the 
n»cklinji: or pouting ( fimlus hisrn.><) mn doubt that thes<' orjrauH are (Mnploye<l fnr the discrimination of 
f(KHi Hubstanccs; but the fact already mentione«l, that the rocklin^ in which thecdfactory oi-^rans had 
been extirpate<l did not take any notice ()f food that was not j»ut close t^» it, jM»int,s to ttuM'onclnsion 
that they are of s*»rvi<'«» oidy in actual contact with the fcMnl itself. 

Hateson ofives also a (;onsidera))lo list of iisluvs which ho has observ(»d to ifot their 
food chiefly by the sense of si^ht, an<l he is doubtless correct in ass(»rtin<:f that the 
majority of tishos bolon*^ to this class. None of si<i^ht-huntin<jf (ishes while living 
in his t^uiks appeared abl(> to see their food by ni*ifht, or even in twiliijfht. None of 
the tishe.s whi<-h ho eniunoratos as belongino^ (o this class showed symptoms of interest 
when th«» jui(*e of food substanc(»s was put into the water, and other evi<lence is 
)>roucr|it forward to show that the of sm<»ll plays little or no part in helpinj^ 
thorn to dis(M)ver their food. 

I have not sttiditnl any of tho species montioni^l by Bateson. but for tho forms 
studied by mo, which have an extensive supply of terminal btids on tho outer skin, 
I fully coniirm most of the stntements quot^^d alK)vo, save that in det(»rminin<if the 
part played by si*^ht I <lid not blind any of my tish(»s and save that the stutement 
that in fishes of his first *ifroup "at no time wjis there the slightest indication of any 
rocomiition of anv food substunce by sij'ht"' is strictlv true of none of mv tishos 
exce]>t AmrfuviiM^ thouj^h in some of tho other cases it is approximat(»ly true. 

Tho only imporbint rosptM't in which my observations are not in harmony with 
those of Batoson is in connection with tho jMirt playcnl by the* sense of t^isto in some 
of tlu'se tv])os of tishos. I have studi(»d tho gustatory reactions of fishes closely 
allied to the rockling and having the same arrangement of terminal buds on the barb- 
lets and pelvic fins, and am convinced thtit Hatosoirs failure to get <*lear gustatory 
reactions from these orgjins was dm* to tho insuflicioncy of his methods of experi- 
ment rather than to tho ab.»<ence of the fun<tion. In ofonenil, it mav bo stated that 
the ])jirt played by th<» gustatory reflex in tlu* case of fishes having an extensive sup- 
ply of terminal btids on the outer skin is of vastly great4T im|M)rt4ince than l^atoson 
ap|x»ars to luivo recogniz(»d. 


The only othrr pupor of im|M)rtan<*o dealing with the sense of taste in the fishes 
ex p<»ri mentally which has come to my notice is the jrreat iiionog-raph on the senses of 
tit^te and smell hy Na<r*d C'*^)- II** investitrsiteil the sense of ta.ste in the following 

(1) Fkesfi-w vTER Types: A nfjn Ufa aufjul/in {o\t\ and (\n'\U' \o\n\g); C!fpnnns< rnrftio; HtirhuA flttria- 

(His; Ia iifisrils rt fJni/iis; ( itistt'mstriis ariittuhis; d'ohnis jinr'mfltis; SilnrtiM ffhtniji J VOllDg 
S|HM"iliH^n ) ; f 'niiit'ijt l^issifi.'i. 

(2) Makink Tv»*k>: Jurist In rn.*; Sit//iitui (^itnlus nwi S. ctiuinda; Sj/iiij)nithti.t /i/i<.v; I'rtnHtsfitpuJS 

tuuifn r; IjOfthnts insnitin'nis. 

Najj^el tested all the fresh-water fishes mentioned in this list by hrintriii^ hitter, 
sour, swe<»t, and salty solutions in conflict with the skin, without jfettin<jf any res|H)nse 
to the stimulus. Thus, the caip, wels (X/7///v/.v), and stickleha<-k did not res|>ond to 
a stimulation of the skin of the hody with ({uinine, thouj^h the last-named fish gsive 
an immediate resjxMis<» when the s<»lution touche<l the lips. He concludes: 

III tho fn'T'li-wattT HhIh's. acmnlinjr to my nbH<*rvati<uiF, tho iMjwer of taj«te is <'ompletely larking 
in the oiitiT skin: <»r, nion» pn^iH-ly, in no i»iirt except tlie hea<l in there piiFtatory sensil>ility. 

For su<-h of these forms as jx)ssess no terminal huds on the skin of the body this 
is doubtless true: but for the other fishes, includinjj, doubtless, Sfft/rftx and ()fpriuus^ 
it is certainly a mistake. In gadoid fishes I j^ot a <'lear reaction aj^inst quinine 
solution when it was applied to the free fin niys, which are known to l>e supplied 
with terminal buds, but not from other parts of the skin. 

Amon^ the elasmol»ninch tishes Na^el found ScylUuin ratuhhs and S, vinilexila 
to Im^ sensitive to very dilute solutions of vanilla all over the bodv and fins. Bitters 
were not p(M'<*eivedthus, nor oil of rosemary, but they are very sensitive to creosote, 
lie controverts S<hwallK»'s arcrument that the terminal buds of the outer skin of tishes 
pr<>l>ably have a fjustatory fun«*tion !)y reason of the similarity of their stru<*ture 
with that of taste !>uds in the mouth, and concludes: 

A n*al S4»ns«* t»f t;istt% >\\v\\ as man an«l many t»thrr animals have in the nioutli, apjiears to Ije 
a)>S4Mit in thr <»nt«T f-kin of all iishfs and Amphihia. 

It will apiwar from the followin<r iwijros that this conclusion is erroneous. I will 
merely add here tiiat if Na«^el had worked with sapi<l solutions, with which his fishes 
were presumably already familiar, instead of with substances like sugar and vanilla, 
toward which no <learlv established reflexes hacl been established in the natural 
environment of the fishes, his conclusions mi«('ht have l>een different. 


The t<*rminal buds of the fishes tabulated above, and doubtless manv others 
which mijjht Ik? mentioned, are of the same type and presumably provided with 
similar innervation bv conununis nerves, for cutaneous bninches of the communis 
root of the facial nerve are known to reach the areas ])rovid4*d with the buds in all 
cases whi<h have been ade<|uately studi(»d. These orjipms may therefore all lie 
defined mori)holo<ifically as !>elon<»'in«i: to the communis system of sense orof:ins, along 
with the taste bu<ls of the UKUith cavity and as distinct fnnn the lateral-line organs 
and all oth«»r ty|>«»s of sense orgsuis. In order to supjxnl this p>sition there remains 
uierely tin* proof that the terminal buds and taste buds have a similar function. 
This evidence is presented subsequently in this |)ii[>er. 

248 HiTLLjrrrx <>k thk iiNrTKii htatfs fish (xiMMisaroN. 

T\w. toniiinal l)inls of tislips hnw 1mh>ii often dpscrittpd mid liffurod, mid I htive 
Iittti> to tidd to tlw ('liissii-iil (li'siTiplions siivo in tho iiiiitt4'r of (lis(ri)iiitioii iiml inner- 
Viition. Ttiosr in tlu> mouth :ii-i' supjtlifil liy limn dies of tli<> x, ix. and VII pnirs of 
cniniill norvcs, tlio lifst Iwo norvos supplying those in tho jrill iTi,'ioiis and th« piv- 
tronmtic hnun'h of the jrlossopliarynfjcus also ninnin<r fonvaid to siip])ly thoMo on 
tho liyoid an-h (toiiiinc). The <'i)niniiuiis ro<it of llie fai-ialis ( - imrlio int4>niiediii of 
liiiniari anatomy) and its f^enii-iilate jrnnjrlion supply the bisfe Imds on tho piihite liy 
tlie r, [iiilatiiius facialis (— <,'re«t suiwrlicial iiitrosal nerve of iiijiii). iind otlipr liiids 
on the liniiijf of tlie i-heek. on the jtiws, and on the lips liy other limni-lKH, some of 
whieh lire stvitiidai-ily associated with bmnehes of the tn^iiiimis and most of wliidi 
have no hoiiiolotrties in miimmalian anatomy, though some one or more (»f tliftin 
pi-olmbly repieseni the chorda tvmi>aiii. 

In A /I"/ III '('■ 1 lia\e shown (o|) that terminal hudsoeciir in the skin of praeticiilly 
the whole IhhIv siirfiwe, most nlnindantly on 
the liari>Iet.s mid dimiriiHhin}r in frefpiency 
Kiwtird the tail. These Intds (see tip. 2) 
rest Oil a low ]NipillH of t)ie dermis, (|uite 
'lilferent fi-om that lipni-cd liy Merke! ('SO. 
lilate V. tifj. 1) for the lerminal Inids of 
s;/iifiM. His lipiire shows a iiLiich smaller 
<irgan, restinfi upon a jfriiitly elongated 
papilla in an epidermis which is ap[>;ireiitly 
Ihicker than in ,!//» /»/•-*«. Merkcl states 
("Sd, ]). 7-2) that terminal liuds always occur 
on Slid) 11 dermal [nipillii. While tins is cer- 
tainly the f,'eneml nrle, we lind (h-cus ion ally 
instances where the |»ipilla is aliseiil, iis on 
the lilliform tins of the hake, where I liixl the 
buds iiiilMsldcd in the epidermis and extend- 
ing only piirtway thnaij.'h it, with a layer of 

Kiiiihi-mis.'.ii ,ii..w|«|.iiiuiin.i..rii.,«n-.«vni. unnnKlJUed epidermal cells JH'tween (he bud 

and the tlermis. 
All iMirtjs <if the l)ody <if Aiiniiini" which an' supplied with ti'rmintd buds am 
reached by briiiiehes of the facial nerve from the peiiiculnte }r;in<>|i[|||. In other 
words, the nmii from the <'<immunis root of the faciiilis are distributed to in^irly the 
whole outer ImhIv surfiiee of this Itsh. On the distal side of the {riin<r|i<iii these niiiii 
usually join themselves t<i otliei' cutaneous branches whi<h are iihyl<ij;enetieally older. 
iM^hmyin;; to the jfeiiend cutimeoiis and latciiddine systems. Kven the preat recur- 
rent branch into the trunk, the ramus lateralis aceessorius. which passes oiit of the 
craniiiui as a pmctieiilly pure counnunis nerve, anastinuoses with the spinal nerves t\\ 
t heir ^intrlia and its liliers are ultimately distributed alont; with lhof;enend cutaneous 
(ilH'i's from thes<> spinal (;miplia. V\^. :> illustrates the couises of the chief cutaneous 
bninches of the eomnmnis system in Aiifiiinix ;«*•/««. the iierves of all other systems 
lieiuff omitted from the sketch. 

Proximally of the jjeniculate (ranglion the communis rofit of the facialis pursues 
mi um-omplicated to the primary gustatory ce.nter within the medulla 
oblongata, lu most lishes this i-oot juisses Iwck close Ui the floor of the fiiurth von- 

THE iir(;an and sense oi.' taste in eishf-s. 249 

Ii-ir)i> II.S till' fiiM-iciilii-i i-<iiiiiiiiinis (—fuse, solitsiriiis <if m;ii>iiit:ils) to toniiiimti' in tlio 
V!i;p'' '"'"' "f •'"' >!">i'' '^l**. !in<l ivi-i'ivt's ill its foiirsc tlio i-Diiiniiinis root of Die jjUk- 
sojilmry II irons ihtvp. Itiit in >iliin>ii1» uiid lyppiiioids, wlicio tlic vorv almndiint 
t<-niiinul liii'ls of llio oiilor skin :ii-i' ull iitixTviiti-d fi-oin thi> (-oiiiiiiunis root of the 
fiifiiil niTvc. tlio crmsi'inii'Tit imicuso in tlio -^izc of tliis rinit has ii'STiltcd in a great 
I'nlarjii'niriit rif tlio t-o|ilii\]ii' ond of llio irii>laIorv contor (vnffal lolio) ivliicli apivarH 
on tlio dorsal -nrfaoo of tho ol.lonjrata as tln> faria! lolio. This slnii-turo is [uiirod in 
silnroid- and was forinorly rjtilod llio !> Irijroniini. an injwlmissil.lo tonn. sinw it 
lias nothing wliatovor to i\n «itli tlio trijroininns norvo. In cyprinoids it is iinjHiired 
and is roforrod to in tlio cdilor litonitiiro as tho tnlx-rciilnni iiiipir. 

Tlio oyprincii.i li-ln- also linvo Ion;: Ih'oii linow n to liavo toniiinal (mils {/W/irr- 
'./-;/-//-) wididy distril.Titod ovor tho cmt.-r IhmIv snrfaoo: l>nt noithor tho innon-ation 
of Ihosp oripins nor tho o\aot i •iiii|>(isitioii of tlio i-ninial norvos lias ovor tioen worked 
out in any . yprinoid lisli. A oiiisoiy oxanilnation of a sonos of s(>otions propuretl 

l.y tho \V.i;r,.rt inoihod throii<rh tho ontlro hoad and IkmIv of a small fjold-fish {Cimi"- 
.■-iiis •iiii--ilfi.-<) has oonvincod mo that tho saino londitions in -.'ononil pivvail in the 
lyprinoiils jis in tho siluroids. That is. tho onoriiioiis size of the vajnil lolws of 
oypriiioid> is oxphiinod liy tho fa<-t thai those arc the terinina) <-entoi-s for the viist 
luHiilH-r- of iiorv.' HIkis ont.rino tho lii-iiiti liy way of the ix and X iiorvos from the 
jKilatal orjran. ihis roiiiarkaldo siniotiiro lM'iii<r orowdod over its entiio extent with 
taslo Inids ;ind sorvin^r to tiltor foiM ]Kirli.|os out of tho iinnl taken into the 

On the iitluT hand, tho luli<'nidiim ini|Kir. i)r fiicinl lohe. roooivos the ontlro 
<-omnuinis rtwit of tho faoial nor\ o. This r<Mit loieives lil)i'r> from pmotioally all 
(Kiits of tho oiilor Mirfa.o of the l.iMfy. :iii<l wo msiy infoi l»y stnalofry with other 
lisho> that tlio>o tilHTs i-onnoot with the torniinal Inids in thoo oiitaneons areas, 
thoiifih we have as yet no aotnul <loinonstnt(ii>n of this faot. The tenninal hnds of 
the skin of tho head ai-e supplied mainly, ns in Am- lur'i". h\ wiiy of tho infniorliital 
trunk. The tenninal hnds in the skin of the body of the jfold-tish are not. however. 


supplied by a ramus lateralis ac^cossorius, or recurrent facial nerve, as in Am/^niruM and 
the jifadoid fishes, for this nerve, as has lon^ Iwen known, is absent in the cyprinoids. 

There is, however, in these fishes an intra(^ranial anastoniasis l)etween the 
v+vii pfanglionie complex and the ix+x i'omplex, the composition of which has 
thus far remained unknown. This proves to !)e the reeurrcMit bninch of the facialis, 
carrying connnunis fil>ers from the j^eniculate fjfan^lion into the trunk. The details 
of the periphenil distribution of these fibers have not l>een fully workcnl out, but the 
main iiath in the j^old-fish is as follows: 

The geniculate ^mglion of the facialis is clearly separable from all other 
jranglionic nuisses of the trigemi no-facial complex and is comix)sed of two iK>rtions, 
c»ach of larji^e size. The more dorsad portion corresjHmds to the greater jmrt of the 
trjinirlion in other teleosts and distri})utes its tibcM-s chi(»flv l)V wav of the infraorbital 
trunk. The more ventnd jK)rtion sends cephalad a very large palatine nerve, and 
caudad a still larger nerve which represents moi'pliologically, though not to|)ograph- 
icallv, the r. recurrens facialis of the siluroids. (»tc., or the facial root of the r. lat(>raiis 
accessorius as found in the cod. 

This nerve i)asses Iwick along the latenil side of the great auditor}^ root and at 
the lev(»l of the superficial origin of the ix nerve it divides into seveml stmnds, one 
of which pisses dorsally of the ix root, the others v(Mitrally. These latter, however, 
jMiss upward so as to lie, farther Iwick, dorsjilly of all of the vagus r<M>ts except that 
of the lat4*mlis branch of the vagus. All of these communis fibers now join them- 
selves to the r. latenilis vagi and, passing through the ganglion of the lattc^r nerve, 
Ijoth components enter the body of the tish l)Ound up in a single nerve trunk in 
whieh the fine conmiunis fibers are for a time completelv surrounded bv the coarse 
latenilis [ib(»rs. The connnunis fibers go oif in successive branches along with 
lat4»nilis fil>ers. The details of the listrilmtion have not been worked out, though I 
think it would not be ditticult to do so with the material at hand. It is highly 
probable that the connnunis fibers are for the terminal buds siwi'sely distribut^^d 
over the skin of th<* lK)dy and that the terminal buds of the trunk are all iimervated 
from these connnunis fibers in the r. lateralis vagi, just as the buds in the skin of 
the head are innervaUnl by other comnuinis fibers from the geniculate gtinglion of 
the fa<'ialis, an arrangement substantially identical in morphological plan with that 
of the siluroid fishes. 

The conditions here, so far as studied, confirm essentially the conjectures to 
which I was led from a study of the lit(Miiture (Ileirick, 'J>J>, p. 40<i), and accord so 
completely with the morphological interpretation there proiK)sed that we merely 
refer the reader to that pjtssage in the Menidia paix^.r. 


The cat-fish (Aifirinnfs n<hitJoAUi<) ujxm which this series of experiments was 
con<luctcHl (except a few ex])eri!nents specificrally designat4»d) were hati^hed in the 
open at (iranville in the spring of VMM, In Octol>er of that sjime year they were 
taken to the lalK)ratorv and kept through the following winter in tanks. Microscopi<» 
examination of the skin and barblet^s shows that their skin and cutaneous sense orrans 


at this age are praeticiilly in tlu* adult condition. Durinor tho winter thoy wore fed 
on various kinds of meat chopped tine, sometimes cooked, hut usually raw. 

In one small a<piarium were kept half a dozen cat Hsh, several ortlinary ''shiners'"* 
(x\W/v/^//.vsp."), and some small ** spotted suckers" {Mnnjirmtn ////^A//*«'/y>/< Kiitines^jue). 
('ju^ual ohservations mad(* during; the winter while feedincj showed that the shiners 
use the eyes chiefly in capturing their foo<l. A hit of mc^at dropjXMl into the water will 
usuallv Im» seized instantiv and devoured hefore it has time to sink to the lK)ttom of 
the tank. After it has fallen to the InUtom it is apt tr) he long overl(K)ked unless the 
tish hapf)ens upon it in its aimless wanderings, or unless its attention is calle<l to it h}' 
the movements of other fish(\*^ which mav l)e eatintf it. These fishes, when ohserved, 
are usually swimming ahout in the midch^pths of the tank, not resting near the 
lK)ttoni. I have ohs<»rved the s:im(» hehavior in Mmiilni and other large-eyed species. 

The l)ehavior of the suckcM's was totally diffident. I'hese fishes lie on the Inittoni 
most of the time unless disturl>ed, though if frightened thev are verv active, swim- 
ming i>owerfully and leaping out of the water. When fcKnl is thrown in they never 
iKiv the sliifhtest attcMition. nor are th(»v attracted hv the sight of other fishes 
strujrirlinir f<>t* the meat. Thev are exceedinglv shv and nirelv eat when under 
okservation. Thev lie ciuietlv nuich of the time or swim slowlv alxmt, dragmng the 
fleshy lips of the highly protrusihle mouth over the hottom of the tank. If they thus 
hapi^en u}X)n a hit of nu^at this is sucked into the mouth, worked ov(*r with tlie 
pharyngeal teeth apjKirently, and then often ejected forcihly from the mouth, to l>e 
agsiin taken, jn^'haps, and the process re])eated— a l)ehavior very chamcteristic of the 
wav thev take the hait. 1 am told hv fishermen. 

ft ft ^ 

The cattish, like flu* suckers, keep strictly to the hottom of the tank. They are 
often quiet in the darkest corners or lying undcM* dehris, hut much f)f the time 
are slowlv dniirjriniT the mental and |K)st-mental harhlets along the lM)ttom. The 
nasjil l)arl)lets are held projecting well upwai'd, and the maxillary Iwirhlets are 
directed outward and backward. theiitip> trailing the hottom or waving gently hack 
and forth. They appear fh^im* to use their eyes directly for cjitching focxl to the 
slightest (h^gn^e undcM* the <*f)nditions of these experimtMits. No attention is pai<l to 
partichv^ of food thrown into the water. (»v(mi though they setth* down within a few 
millimeters of tlie nose or harl)let of the fish. Th«» on I v case ohserve<l hv me in 

ft ft 

which th«» eyes seem to serve* in finding f(K)d is when a large piece of meat is thrown 
in and one tish In^jrins to •"worrv" it. His movements mav attract others until as 

^^ ft • 

manv fish as cjin reach it are all tugging at it at once. If, however, a shadow is 
caused to fall upon the water, as hy hovering the hand over the aquarimn, the fishes 
an* irreatlv disturhe<l and dart wildlv ahout. Thev alwavs seek the darkest corners 

^^ ft ft ft • 

of the tank and lie under dead leaves resting on the l)ottom of the tank for the most 
part, showinir that the eves are not hv anv means functionless and the fishes are 

I ^^ • ft ft 

strongly negatively phf)totactic. 

If the cat-fi>hes in the course of their aimless movements along the floor of the 
a<|uarium touch a hit of meat with tin* lips or harhlets, it is instantly seized and swal- 
lowed. Food in the inunediate neighl>orh<KMl of the fish is not discovered at cmce, 
hut after a timeap|x^ars to affect the fish in some way, prohahly through the sense of 

'i.V"/ro/iix hii.M vcn simill tu)M'n*iihim inipar aikI vacail IoKm-s. tin* laltfr S4-un-<-ly lander thnii \u tUo tttii, Mtiiidia. &wi 
phvMfir'listonM fishes gononUly. Fnmi thif one may safely infer that cutaneous terminal buds are not as highly developed 
in thi» form as in the largt^r cyprinoiils. 


smell, as the inaxillary harblets hocr'm to wave about nK)re actively and tiiiall}' the fish 
becomes restless. He does not find the food, however, unless in the course of his 
mov^'*ments it actually touches some part of the l)ody. 

Durint^ May and June, 1J)()2, more systematic exi)eriments were undertaken with 
these fish, and since these experiments are typical of those subsecpiently jMM'fonned 
on other species of fishes 1 shall recount them in some detail. At first a few s[MH*i- 
mens were t4iken out in a shallow tray and the attempt made to feed th(»m in various 
ways under close observation. They were, however, so nuich frightened by the 
exix)suie to brit^ht dayli^jflit and l)y the proximit}' of the observer, in spite of all pre- 
cautions, that no reactions could beolitained which were at all satisfactory. A bit of 
fresh meat on a loncf- handled needle could be thrust slowlv toward the fish as he lav 
(juietly on the l)ottom, rub])ed over his Ixxly or on the bar})leb^, and even over the 
lips, without evoking a movement of any kind in response. The same observation 
was made with the spotted suckers. The fishes in l)otli cjises had been without food 
for several days and were vei'y hungry, but were obviously too much frightened to 
respond to the food stimulus. 

On another occasion the same conditions wmm'c prepannl, except that a few dead 
leaves were littennl over the bottom of the tray. The fish when placed in the tmy 
immediately sought the shelter of the* l(*aves, and, after a suitable interval to enable 
th(»m to become accustomed to th(^ |)lace, the feeding experiments were rejwated. 
Selecting a fish which was entii'cly concealed undcM* a largi^ leaf, save for a projecting 
bai'blet, a bit of meat on a slender win* was gently passed down into the wat(»r in 
such a way as to touch the pi'ojecting barblet. It was instantly seized and swallowed. 
This was n»p(»ate(l many times with s(»veral of the fishes. 

In subse<|uent experiments the fish were not removed from their own tank, Init 
the water was drawn ofi' so that it was only about six inches deep. Here they would 
lie inider the leaves and the experiment <*()uld be continued with a mininuun of 
disturbance to the fishes. The experiment of touching the barldet with meat was 
repeated hundreds of times with an almost invariable result that the fish instantly 
turned and snapped uj) the morsel. If the meat w^as nu^ndy held ver}- close to the 
i)arbl(»t it usually produce<l no nvsponse. The reaction was ol)taine(l ecjually well, no 
matter which barblet was touched. 

In a later series of experiuKMits I foimd that the fish would almost always turn 
and s(»ize the meat if he were touchcMl at any point on the h(»ad or body. If the tail 
of the fish proje(;t(»d out from under a leaf and th(* skin n(»ar the root of a tail fin 
w^en* touched with m(*at tin* fish would tui'n and seize* the nn^at. This reaction was 
not so unifonnlv made at first as that from tin* ])arblets, but after a dozcMi or so of 
trials it followc^l with ecpial promptness and uniformity, the fish apparently re<|uir- 
ing a little practice* to l(*arn the mov(*m(*nt p<*rfectly. 

The experiments last described wen* rep(*ated th«* next day and by this time it 
wits foimd that the fishes had become* so tame that th(*v would take the meat if 
offered to th(*m in the open, without the shelter of the d(*ad leaves, though not so 
certaiidy as when under the cover of the l(»aves, often taking fright from the shadow 
of the oliserver's hand or from some other cause. 

In none of thes(* cjtses did the fishes appc^ar to see the Imit or to i)erc(*ive it in any 
wa\' other than by actual conbict with the skin at some ix)int. If the bait were held 


a moment in fi*ont of them and then moved .slowlv awav thev would not follow it. 

• % * 

If, however, it touchi^l a Inirhlet and tlien moved rapidly away l>efore the li-sh had 
time to seize it, then the tish would sometimes follow it a short distanee. 

At this \K)\ut the relations of vision and smell to these reactions should receive 
some further consideration. These younjr li>he>, like their adults, sjkmuI iimch of 
their time huried under the debris of the bottom, with iwrhap^ a harlilet or a }K>r- 
tion of the tail only projectintr. liuler the>e cinumstances it is easy to apply the 
stinudus to various part> of the skin with the assurance that the contact is wholly 
invisible to the tish. Manv such e\'i)i*riment> show deci>ivelv that the reaction takes 
pla<-e in the sjime way whetiier the tish is able to see the stinudu> apj>lied or not. 
The visual factor In'inir >o conclu>ivelv ruled out, 1 have not thoui'ht it necessarv to 
blind the tish for further control. 

This conclusion of cour>e imi>t Ih' limited >trictly to li^h of the sjKvies and ii^e 
under i n vest ijrat ion. It bv no means follows that thev mav not subsenuentlv learn 
to use their i*yes in tindin<r fo<Hl. a*^ well as in escapin<r from their enemies. Indeed, 
durin*r the later ex|)eriments of this >eries, after the tishes had been fe<l for several 
weeks almost dailv with meat on the end of a wire, I siiw >ome slitrht evidence that 
thev took note of the bait bv the >ense of sit^ht, but the observations were in no cit*^ 
conclusive. Whether the adult Ain^im-ns nrhnlnsn.s ever uses the eyes in the <-apture 
of food I have no definite information, thouifh from tln^ habit of sj)ending much of 
the time durint^ the day completely buried in the nuid and of feedin<^ chiefly at night 
it is very improl>able that they do so. With the channel <'at-tish, lrtnJt(rn.s^ the case 
is certainlv ditferent. 

Mr. 1. A. Field tells me that while tishinif for b:iss in the HIack River. Ohio, he 
has sometimes caut^'ht lar^a* specimens of Irfuhirns with live minnow.^ as bait. The 
current wa> swift and the minnow> were kept olf the lK)ttom of the river and in 
motion all the time. At the meetin*:^ of the American Association for the Advance- 
ment of Science, at Pittsbur«r< 'July 1, n*'»*J, in the cour>e of a l)rit'f rcjHirt upon these 
experiments, 1 asked the question whether anyone evt'i' cau*rht a cattish on a sjkx>ii 
hook. Dr. L. L. Dvche stated that he ha> occasionally caujrht the channel cat {Irta- 
hir*if<) on a sjK)on in a small laki*, but only in bri*rht >unlitrht. Dr. Ki<renmann stated 
that Ltahinis has nuich I setter evi's than Anniiirns. Thev are not onlv larjrer, but 
the retinal pattern is more nearly like that of otlu'r fishes, while that of Aiiti!»ir*(.s 
is decidedly degenerate. 

The part played by the of Muell is nmch more ditlicult to determine. As 
intimated alMJve, I have evidt'nce that the gustatory organs of the skin can function 
onlv in contact with the >;ii)id .sul)stanct'. The most hiirhlv flavored fcHHl can be held 
within a millimeter or two of the Iwrblct or lip> without calling forth tin* chanu-ter- 
istic instantaneous reflex. I will narrate one t*x|M»rience which was many timci? 
rejK'ated in a variety of moditicjition>. Thn'c li>he> wt're lying (juietly under a small 
water-soaked leaf. A bit of rather stale beefsteak, with a strong cnlor, wa> held on 
the tip of a Hue wire over the edge of the leaf under which they were lying and sejMi- 
rated bv a centimeter or two from the no>tril> of tln' ti>hes. The leaf wjts consider- 
ablv corriHled bv det-av, and doubtless the odor coulil freelv i)ermeate it, thouirh it 
was nearly or ipiite opaque. After some ten M»cond> the flshes l»eg;in to move rest- 
lessly alx>ut in circles under the leaf, which was soon swept away by their movements. 


As a ruhi the lislios swam in ujiitow c'ircl(»s close to the bottom and for a lon^ time 
failed to find the meat, thoufjh thev seemed to l)e aware of its j^eiieral jM)sitioii for 
thev never circled far awav. If the meat were very slowly moved across the a(iua- 
rium the fish could he drawn in this way after it for a considemlile distance, though 
the meat was never found unless in the course of their apparently aindess movements 
one of the fishes came in cont^ict w^th it, when it wiis instantly snapi>ed up. 

This aimless circlinj^ movement may 1)C termed provisionally the KctkhK/ reac- 
tion, since it is so ditTcrent from the chanicteristic movement made when tlu» stimulus 
is in <'ont4ict with the l)ody a shari> turn of the hixly and instantaneous seizinif of 
the 1)siit — which I shall i(tv\\\ i\\i\ tjH^sd it nnj n art ion. Unfortunately, I have not had 
opportunity as yet to carry out extirimtion exjwriments on Aimtunifi to det(M'mine 
decisively the jMirt played by tht* olfactory or<^an in this reaction. (Compare the 
expt^riments on the tomcod narrated Indow.) 

The fishes ujHin which these ex|H'riments were jH'rformed have unfortunately 
lK»en lost. At the present time 1 have a fresh lot of Aim lurnsivy under observation, 
and have already verified many of the conclusions nniched with the first lot. But 
this .second colhH'tion of fish(»s has not, at the time when this report is submitted, 
been in captivity lon*^ enou<i^h to Unome sufiiciently accustomed to their new sur- 
roundinj^s to feed frei'ly and ft^arlessly. After some months of further prelimi- 
nary obs(»rvation, I hope to carry on experiments which may shed some li*j^ht on the 
sense of smell in th(\se fishes. But this nuist be reserved for a latter rejxirt. A few 
subsecpient observations are noted on pajj^es 27(>-271. 

AVe nmst content ourselves at the present time, then, with the inference that 
the sense of smell plays at least a small part in these reactions, for the animals 
In^came slightly restless in the proximity of the stimulus, thou<!fh they were not in 
contact with it; this, however, appears never to provoke a definite reaction of seiz- 
in<( the food, l)ut merely a va^ue reaction in search of food. On th(» other hand, 
physical contact with the irritatin*^ sulistance causes a definite and precise r(»action 
which is pnictically constant. This points either to touch or to t^iste. 

To the relative part played by stimulation of these two sets of sense or*^ans, 
the followinj^ series of cxix'rimentHS was pi»rformed. A half dozen fish in an aqua- 
rium wen* tested a scon* of times with fresh meat on the tip of a wire, as in the 
previous cases. The reaction was obtained uniformly, no matter what part of th«» 
ImkIv or head was touched. Half an hour after the clos«» of these ex|H'rim(»nts a bit 
of cotton wool was wound around the tip of a wire and the fishes were tested with 
this exactly as tlK'V had been with the meat. For the first six trials the barbh^ts 
otdy were touched. 'V\\i\ fish in each case turned and seized the cotton as promptly 
as the meat had been taken. The cotton would be immediately (lro|)pt»d. After a 
few more trials the fishes would «i:i»nerally turn when touched, but would check their 
moveuHMit before the cotton was actually taken into the mouth. Several spe<imens 
were now tested on the trunk with the cotton. One or two turned completely around 
and took the cotton, ))ut generally there was a slight movement oidy toward the 
cotton, which was checked before the cotton was reachi'd. Aft(»r a fi'w furthei- 
tests, the fishes would usually pay no attention to a contact with the cotton on the 
skin of the body and the reaction by the bar)>lets became uncertain, until finally the 
cotton could Iw freely rubbed over the barblets or lips of some of the individuals 
without producing any response. 


These ex|>erimeiit> were uuuiy times i*e|K*aleil. ?soiiietiineb usinj( white cotton, 
sometimes red cotton, and sometimes fresh meat. Tlie reaction was uniformly obtained 
with the meat. If at the ch>se of a few ex}x?riments with the meat a minute pledget 
of cotton was suh*^tituted for the meat, there was feeble or no resi)onse from rubbinj^ 
the Ixxly with the cotton, thou<irh uiH)n touchintr tlie Imrblets the tish would usually 
turn and often would seize the cotton and drop it agiiin at once. After several re}>e- 
titions, the tish I Kvame wholly inditTerent to the cotton, no matter how it was applied, 
or they would if touched upon a l>arblet turn toward it without ))itingit. They were 
now again tested with l>it> of meat. This they took as eagerly and as precisely as 
l>efore, showing that they were >till hiuigrv. 

After the interval of a day or two the lishe> would still apjK'ar to rememlKM* the 
cotton, and 1 nirely, after the tirst trial>, got a prompt •*gu>tatory" refl(*x with the 
cotton. If they noticed it at all, they would turn slowly and touch it with the lif>sor 
a Imrblet in a tentative* or in<|uiring manner, only to turn away again without taking 
it into the mouth. This (lelilK.M*at4' movement may lie de>ignat^'d, for re;isons to appear the ttirtlJr nlf<.i\ as distinguished from the instant s4Mzingof food, the 
' * gustator V reflex . '* 

These exjXM'iments seem to show that in the reactions to the meat, lK)th from the 
l>arblet and from the skin of the iKxly. the senses of tiiste and touch lx>th luirticijwte. 
This is in accord with the known innervation of the skin and Imrblets, for all j)arts 
of the IkkIv surface receive general cutaneous (tactile) nerves, and all pjirts are plenti- 
fully provided with terminal buds (taste buds) which are innervated by communis 
(gustiitory) nerves. The exi)eriments further suggest that thesi* two sensory factors 
can be exix^rimentally isolated by training. 

The fishes having ))ecome accustomed l>v brief tniining to make the simple reflex 
of seizing the food under the stinmlus applied to any j^art of the Ijarblets or skin, 
and doubtles*^ utilizing l)oth gustatorv and tactile sensiitions, the gustatorv factor is 
eliminated bv the su))stitution of cotton wool for the meat. The tactile sensation 
alon(» proves to b«' sullicient to set off the reflex after the training previously given. 
The stimulus is, however, never followed by >;itisfaction and i> soon given up, the 
fishes after further pnictice not reacting to the tactile stinmlus alone. If, howt»ver, 
tlu' triistatorv >ens;ition i> added, bv the .Nub>tituti(»n of meat for the c(»tton, the 
original reflex i> given a^ promptly a> befon'. This would seem to indicate that, 
while the tactile si»n.N;ition alone is not sufficient to maintain the n»flex, the addition 
of the ifustatorv eh*uu'nt is sufticitMit, and tlaM'efore that the gustatorv elein«*nt is 
the es.>ential element in setting off the reflex. This hy}>othesis wa> tested by an 
extensive series of exjK'riments similar in plan to those last desi-ril>ed. 

In genenil there wa> no noticeable ditference Ix'tween the reaction to the white 
cotton and that to the red, though in some caM's, es|)ecially toward the end of the 
series of exix*riment>, after the fishes had learneil to pay no attention to white cotton 
wlien touched at any point by it. they would sometimes turn and touch the red 
cotton w ith the lips or a barblet, immediately to turn away agam without biting the 
cotton as they did at flr^^t. The reaction is not the quick turn and instant seizing of 
the Iniit, which I have termed the ''gustatorv reaction,*' ]>ut a more delil>emte move 
ment similar to what I termed alnive the "tactile reaction.*' This occurred only 
when the cotton was in plain view at the time of the contact and is probably in this 


ctist*. partly a visual re.s|)oiise, called forth by the similar apiK^arance of the red cotton 
and hits of JKiefsteak on which they were habitually fed. It was not by any means 
constant, for, in j^eneral, after the tirst few days, contact with neither color of cotton 
calh^d forth any response whatever. 

After this result was reac^hed, I dipju'd the pledgets of white cotton in the 
filtered juice of fresh beef and touched the body surfaces and barblets with them in 
the siune way as before. In all eases I j^ot a tv])ical '^ gustatory '' reaction cxai'tly 
the sanu' as with the meat, and this n^action i>ersisted aftiM* many trials with no 
diminution. The cotton was tiiken instantly into the mouth and tugged vigorously. 
No amount of training served to eiiKlicate or to weaken this rellex. 

I next preinmrd a small bulb syringe, with the delivery tube drawn out to a 
very line point. This was *illed with the water in which the lishes were and a tine 
jet dinu'ted against their bodies. They either paid no attention or were disturljed 
and swam away. I now substitutcid for the water in the syringe^ the juice of mw 
beef pressed out and strained. When a jet of this fluid wasdiiected against the side 
of tin* bodv, the fish alwavs instantlv turned and trit^d to tak(» tlu' end of the svrinjjfe. 
The reaction was identical with that produced when a corn^sponding part of the body 
is touched with mw meat. 1 invariably got the reaction, both from the sides of the 
bodv as far back as the root of the tail tin and from the skin of the head and barblets. 

1 also tested the fishes with bits of r(»d brick held in forct»ps. The forceps 
seemed to frighten the fishes. They either paid no attention to the contact with the 
brick (when touched in such a way that they could not see the point of contact), or 
else the harsh contact set^ned to frighten them. I then touched them on various 
parts of the body and the barldets with bits of }»rick which had been soaked in raw 
meat juice. In most cases they would turn and touch the brick with tht», lips or take 
it into the mouth, }>ut often they seemed frightened and would .swim away. I then 
gave* them a few bits of meat with th(» forceps and found that th(»y took it eagerly, 
])eing very hungry, ))ut it had to Im» given more cautiously than with the wire, as 
they were afraid of the forceps if th(\v saw them clearly. 

Next I dropped bits of brick which had l)een soaked in meat juice in front of the 
fishes as they lay under leaves with the barblets projecting bt^yond the edges of tin* 
leaves. In all such cases, upon touching the brick with a barblet, they seized the 
luick and bit at it viciously. Often they wouhl return to it a second or third timr 
and try to bite it. 1 dropped similar bits of brick which had not been soaked in 
meat juice in front of tluMu in the same way, l)ut they paid no atti'ntion to them, or 
in a few cases they would touch them with the barblets and then swjni away again 
(•'tactile" reaction). They never attempted to bite them. C'hnirly they taste the 
meat juice in the bricks when they ure touched by a lKirl)let, and the experiment 
when the bodv was touched 1)V a similar brick held in forceps shows that thev taste 
the juice by the body also. 

On one occasion 1 tested the. fish(\s with pieces of cooked meat that had been 
long lu^iled so that nearly all of the extnictives were drawn out. The experiments 
were conducted jwst like those with the raw m«'at, but the fishes gave by no means 
so clear reactions to it. Upon touching the sides of the body, the fishes usually paid 
no attention to the stinudus, treating it just as they did cotton. I then touchr'd the 
Ittirblets a few times, and to this they would gi»nerally react by turning and Uiking 


tlu; meat, but not always nor .mj promptly a.> with fre>h meat. l'ix>n testinjif the 
sides of the UkIv a*r»ii»» after this I'Xperienee 1 j^ot a reaetion. The H>he> would turn 
and toueh the meat with the harhlet or lip> lK»fore takinjr it, rarely <rivin<^ the (piiek 
reaetion ehaniet4*ristie of fre>h meat. Evidently tlie eooked meat hits less tit^te to 
the tishes than fresh meat and tlii< interf«*re> with the reaetion. Thev eat the eooked 
meat when thev are sure that it i> edihle. 

The>e ex|)eriments, all of whieh were many timo reju'ated and eontrolled, 1 
think show eonelu>ively that praetiially the wlu>le lutaneous surfa<e i>f Ain'i»irti.s is 
st»nsitive to Injth taetile and i5u>tatorv >timuli. and that the latter eall forth eharae- 
teristie reflexes whieh aie of th<' ^rreati'^Nt value to thf li>h in proeurin<^ foiKl. The 
ti>h normallv reaets to eonta(l> on the IhkIv I»v lH)th tviK'> of stinudi - to the mere 
taetile stinudus (if at all) hy a ti'iitative movement caleulati'd to hriuif the doulitful 
suhstunt'e into contaet with the more hi^^hly .sensitive harhlet.^ or li|is, hut to the 
ta<*tile stinudus aeeompanie<l hy the }^»'ustatorv hy an immediate, nipid. and preei.>.e 
movement ealeulatinl to s«'ize the foiwl. This latter reflex is unvarying and is very 
IH»i*sistent under a trn^at variety of forms of stinudation. The former (" tiictile'') 
reflex is less stable, and mav Ik* rea<lilv eliminat<'d hv a >imple et»ur>e of tniiiiinir. 
Clciirly the jjfustatory element of the >en.sjition complex resultinj^ from a ronta^'t with 
a sapid suhstanre is mon' iminirtant than the* tactih' clement. 

It is clear that in order to call forth the iharacteristic "j^ustiitory " reflex the 
stinudu> nuist 1m' quit^* stronjr and rather sharply locjdized. For when there is only 
a small amount of meat juicr diffused throuifh the water, as by the presence of a 
piece of fresh meat near flu* fish, he i> not able to lo<*;ilize it accurately, but exhiljits 
onlv the **s(»ekinir reaction." I have not a> vet Ik'cu able to convince mvself whether 
the fish could accuratelv Unalize a stron<r and >liari)lv locjdized tru>tatorv stiuudu> 
with no ta<tile element. In all the exiH*riment> in which meat jui<*e w:t< directt'd 
ajTiihi^t the iMnly with a pipette or >yrin»i-e there wa»N dou)>tle» some tactile effect 
pHnluced by tlu' imi>aet of the jet. We know fi'om tln' exiM-riments that pure taetile 
stiumli can U* a<'eurat*'ly localized on the >kin, an<l there can be no doubt that under 
normal condition^ the>e a>sist in the loealization of the f<MKl objci-t. C'omjian* the 
further di>eus>ion in the -\d(h'ndum. pa^'"e> *jTn-271. 


The preiedintr <'X|M'riment> wi'i'c all c-ariied n\\ in the z^Mihij^ical lalK>nitorv of 
DeiHson Iniversitv: the exiH'rimi'nts on marine fishe^ whieh follow wvw made durin«>' 
the sunuuer of l'.H»2 ut the I'. S. Ki>h C'onunission laUinitorv at \V(kk1s IIoK'. The 
feedinjr reactions of thret' tvjHv^s of <radoid> wt're studied, viz, youn*^ |N»ll<N-k (/'///</- 
rhtins /'//v //.v), alH)ut lo cm. loniri hake {i rnftln/*i.s A /////.s). alH)ut 2n cm. lon<^. and 

yOUn<r adult tomC(Hl (Mn'lUHjtuhix Iniitrml). 

As is well known, the hake and tomccnl have a mental barblet whieh i> known to 
lie abuuilantly set with terminal buds and whiili receives lN>th communis and ^en(»nd 
cutaneous innervation. In all three tyjH-s the lips are freely supplied with ternnnal 
bud.^ an<l there is a re<urrent bninch of the facial ner\ c. the ninms lateiiilis a<*ces- 
sorius, which carries comuuuiis filn'rs into the trunk to supply terminal buds found 
on the fins, es|M*cially the free rays of the vi'utral or pelvic fins. The>e fins are far 
forward under the throat. In the iK>llock they are but little modified; in the touK-od 

Y C. B IWH— 17 


two niys are iil)out twiro as lon*^ as the others and for alK)ut lialf th<Mr lenjicth they 
project freely below the rest of the tin. In the hake all of the ray> of this tin are 
suppressed save these nioditied free niys, so that the tin is tilliforiu. hranehed at the 
end. Microst'opie examination shows that the terminal huds are more ahundant on 
the mon* hiijfhlv moditied tins. The* hak<» also has a fret* tilament on the dorsal tin 
prodih!ed hy tlu» (Extension of the third and fourth niys beyond the others. I have 
not examined this free tilament mieroscopically, hut know that it receivt*s conununis 
tihers from the r. latemlis aecessorius. and hav(» no doubt that it also luis numerous 
terminal buds, as the expt^riments show it to bo very s(Misitive to gustatory stimuli. 
The pollock have very lar*^e eyes and are excellent visujdizers. When food is thrown 
into the water, they dart for it and in ^iMieral they take their f(K)d by the visual reflex. 
So keen is the vision that it would ])e ditticult to carry on any exiMM'iment.N. such asi 
have done with the other two species, without lirst blindin<i: the lish. Nor do they 
habitually dni^ the bottom with the free ventral tin niys as the others do. 1 have, 
therefore, not (h»voted nuich attention to this species, preferrinj^ to study more care- 
fully those s|MH'ies in which the <rustatorv n»tlex plays the j^reater ])art in the life of 
the tish. 

Tin hah't (( mphifcis ftmifM), Tlu^sc ti>he>. like the tomc(»ds. readilv adapt 
themstdves to life* in caiJtivity, and are <»asily cxiH»rimented uiH)n in small tanks. 
They are excidlent visuallzers, thouj^h not so nmch so as the pollock. When bits of 
meat are thrown int<» the water thev usuallv catch them before thev fall to the 
bottom, and their ketMi vision nnikes ditlicult such c\p(*riments as I carried on with 
the cat -fishes. They do n(»t si'cm to re<-o<^ni/e by sii^*-!!! food lyin*^** on the bottom, but 
only wla»n it i> in motion. \\\\{ bits of meat, tish, or clam lyin*:' on the bottom are 
usually found by the aid of the fret' ventral tins. Tln'se tishes spend much of their 
time in slowly swinmiin*^ in Jin appan^ntly aindess manner close to the bottom of 
their tank. Durin*:^ these movements the lilanuMitous pelvic tins are >o Indd that 
their tips dra^ the bottom. These tin rays are <|uite Ion*;, and they are usuallv 
directed obli<piely forward, outward, and downward, with tin' two branches of each 
tin widely divari<*ated, so that the four tips touch the »rround in a line tnmsverse to 
the body axis at about the level of the mental barldet. In this way the bottom under 
the tish and for a short distance on either side is thorou«rhly explored a.N the tish 
swims over it, and all food i)articles with which the barblet or free tin ravs come in 
contact are taken l>y a quick and preci>e movement similar to that set oH* in the 
siluroids hy contact with their Imrblets. Hit> of meat or clam on the end of a sleneU'r 
wire could be laid on tlu' bottom of the lank an<l tluMi slowly moved up under or 
behind the tish and the I'etlcx from the ventral tins tested in this way. Such exjMM'i- 
ments, however, had to b(» made with ^reat caution and many times repeated to rule 
out iH)ssible visual sensations which likewise call forth an innnediate rell(»x. 

Hateson ('1*0, /^) records similar reactions with the rocklino { MofJ/t/)^ a vtikUM 
tish with the same jrenenil structure and distribution of terminal buds as th(* hake, 
but with better developed barblets. (On the structure of the pelvic tins of Mnf^lhi 
compare Hat4\son\s account on p. 214 with that on p. ^1?>\ of the same volume.) 
Bateson. moreover, j^ot the stune reflex with tishes which had been blinded, and I have 
not thoQj^ht it necessary to repeat this ex|M'riment, for my tishes »rive sutlici«'ntlv 
clear evideuce that thi« reflex from the tins is wholly independent of vision. We 


have, however, to investi<rrtte tht* pjirt^ played l)v taetih\ trustatorv. and olfactory 

HatesonVs remarks ('l»o. //. p. 214) in thi.> connection on the rocklin*^ may tw* 
(juoted here. The three-ln^arded and the five-bearded r(K*klin*r aie nocturnal and lie 
still all day. 

<ifinTally, ]M)tli the animals takt* iu» iiotirt* <»f f«MMl until it has lain in the water some minute**, 
when they start <»ff in st^inh <»i it. The n»cklin}j: searches hy si'ttin^; its tilamentous pelvic fins at 
rijfht anjrles to tlie InMly. and tlun >\vimminj: alK*ut iVvlini: with them. If the tins touch a piece of 
fisli or otlier s«>lt UkIv, the nnklinL' turn.- it< h^ail rnun«l an<l snaps it up witli ^reat quickness. It 
will even turn roun«l ami examine inn-atahle sul stances, as irla^^s, etc., which <-ome in contact with its 
tins, an<l wliich pre>nmal»ly .<t'fm !•• it tn riMjuin* an ex|tlanatinn. Tin* r<Kkling>* have ^reat j>owerH 
of s<'ent. ami will set «»ff in .H*arch •»! mt*at hi<hlen in a lM)ttlc Huik in the water. Morefiver. a blind 
r«x*klinjr will hunt for its i'^mmI an<l liml it as t^asily as an uninjure<l one. 

The alH)ve, taken in connection with other i>jissage>. shows that this author con- 
siders that the food i> found larjrdv hv scent, and that the tin reaction is essentially 
tactile, thou*^h he has seen the .-itMise orpins on tlie p«dvic tins and recognized their 
resemlilance to tast<* huds. 

Kxamination of stomach contents >how> that thi' normal food of hake is 
largely crustaceans, jmrticidarly .shrimps. 1 titted up a tank with some .seaweed and 
put into it a largt* niunlxM* of prawns (Ptihunnintts), mostly living, Imt some dead. 
UiH)n putting the hake into this tank, they immediately ate some of the dead prawns 
from the bottom and afterwards caught the live ones, Init verv slowlv and with manv 
failures. Tht» res|K)nM' >eems to l)e wholly visual. The tishes would rej^eatedl}' 
jMi.'^s directly over living prawns, touching them with the tins or Inking brushed by 
their antenna', but so long a> the cru>taceans were (juiet th(\v seemed not to notice 
them. If, however, a prawn was killed and crushed and thrown Imck into the water, 
it was immediately found. I'pon another occasion I put a live clam into the tank 
with the hake, v here it remained for several day>, with siphons greatly extended. 
The tishe> repeatedly bru.-^hed over this si])hon with their free tins, })ut never jjaid 
any attention to it, though if a similar siphon were cut otT from a live clam, so 
as to allow some of the juices to escape, it would be immediately taken and eateu. 
Evidentiv lixe food is riot <learlv located l)V the c^ustatorv or<nins of the tins. 

Beside-* obsciving as fully as po>sible the normal feeding habits of the hake, I 
ex|MMimented upon the reactions to stinuili applied to both the pelvic and the 
tilamentous dorsal tins. As mentioned bv Hateson, the ])elvic tins are fn^elv used to 
«»xplorc all manner of substances w Inch may attract the notice of the, whether 
edible or not. After tiiesi' tishes have bei-onii' accustomed to lK?ing fed small bits 
of meat or clam i>r mussel {Mn(/inhi) in their tank, thev inujiediatelv swim toward anv 
small unfamiliar lM)dy with the |>4'lvic tins thrust forward to touch it before the 
mouth n^aches it. Sometimes the tips of these tins close over it with a movement 
stronirlv sujrifestive of irni^pinof. thouirh of course this the\ can not do. 

Fpon testing by contact with meat or other bait, the free dorsal tilament i.s 
found to Ik' quite as sensitive to gustatory stinudi as the tilamentous ventnils. The 
reflex in this case is very chara«*teristic and constant — the fish upnin touching a 
.savorv morsel checks its forward movement and immediatelv '* l)acks water" so as to 
revcrjie the movement of the lx>dy until the object is directly al>ove the mouth, when 


it is taken at oiuc. This retlcx usually (tliou^^h not so invariably) follows ii contiu^t 
of meat uiK>n any jxirt of the dorsal tin, a> well as tlie free lilanient. The reflex 
rarely fails when any one of the tilainentous tins is touched hy freshly cut meat. 
After meat has been in the water for fifteen mimites or more it seems to lost^ its 
savor and the tins may he repeatedly dni*i^<i:ed over it without eallinti' forth u 
resi)onse, and the sjime is true of the barhlet and lips. 

1 tested the filamentous fins with a wisp of cotton wool on a fine wire, as I did 
the eat-fishes. It was rarely noticed at all by the pelvic tins, but at the first contact 
with the filamentous dorsal the fish reaet4»d just as he did to meat with which he had 
been teste<l inuuediately In^fore. U|K)n repetition, tlie response was soon discon- 
tinued. Foi* a few test-s the fish would pause, and perha|)s Iwu-k up slowly so as to 
smell the suspiciou.s object or touch it with the barbh^t, IkU it was not taken into the 
mouth. After from two to ten tests no further attention was paid to the cotton, or 
the fish w<mld pause a monu'nt without backintr up. This ex|H'riment wa.*< many times 
repeated in the cours<» of the first day of its trial and daily then»after for Nonie time. 
If thrive or four hours intervened between two seri(»s of about twt»ntv tests, the first 
one or two tl^sts of the second series mi^ht be followed by an incomplete reaction, 
but after that usually no notice was taken of the cotton. The fi>hes apparently 
remembered the precedinj:^ testes. Hut if more than twenty-four hours intervened 
between tests, the process of tniininjif usually had to be ^one over a<^ain. 

The fact that the hake does not a|)pear to rememb(»r the dillVirence between the 
pure bictile stimulus and the tactile plus the i^ustatory for so lonjx a time as the cat- 
tish does is prolwibly to be ex])lained by the fact that the number of taste buds on 
the filamentous fins of the hake is nuu*h less than that on the i)arblets of the cat-fish, 
and therefore the jifustatory (*lement in the sensation complex is doubth^ss nmch less 
in the hake. The wdiole course of the cxperiment> indicates that thc» response is in 
fact nmch more strontjlv tactile in the* hake. 

I)urin<( the course of these experiments I oft^'u alt4'rnated bit>j of meat with the 
cotton wool, and at other times su))stituted cotton that had U'cn soaked in <-lam juice. 
In Unvse cases 1 always j^ot tlie chanu't^'ristic <j^ustatorv n^action by all of the fihimen- 
tous fins, no difference bc*inj^ observai)le l»etw(»en tin* reaction to meat of <*lams or fish 
and that to cotton soaked in filtered clam juice. 

1 also tested the hake with i^'-elatin which had been soaked up in cold water. 
Shreds of the w^ell -softened *relatin were fastened to the end of a wire and brou*i^ht 
into <'ontactwith the ImkIv >urface. The reactions were identical with those obtained 
with whit^' cotton. The ifclatin shreds are verv nearlv coK>rless and absolutelv 
tasteh»ss to my tongue. Hut to the sense of touch they are almost exactly the sann* 
jis the bit> of fresh clam meat with which most of these ex])eriments have been <-on- 
duct^nl. The hake at first would take the l)ait when thi» filamentous dorsal was 
touched, Imt if the jrelatin was taken into the mouth it would be immediately 
rejected, and after a few^ trials the fish would no lon^n'r resiH)n(l to the stinndus. 
He acted in the sjune way when the iKilvic fins were stinmlated. Shreds of the 
softencil *jfelatin fallinj^ through t\w. water were somc^times noticed, but nirely taken 
into the mouth, and if so, were inuuediatelv rejected. Similar shreds IvinL' on t\w 

ft ^ ft c"^ 

bottom were netrlected, even thou*^h the bariiiet and filamentous tins dra<'*<'*cd over 
them repeatedly. 


I next took small clam sholls that had l>oen lyin^ lotifj in the tanks containing 
the tish and wciv thorouorhlv ch^anod of fleshy matter and which the tishes had not 
jxiid any attention to for (Liys. These I dried and warmed and then tilled 
with meltinl jj^elatin which had l»een previously softened up in cold water. Upon 
coolincr thcM-e results a mass, colorless, tit^teless, and odorless, which feels almost 
exactlv like the flesh of the clam, which has often l)een fed to the tishes in this wav. 
r|K)n droppincr these shells into the water, th(^ tishes eajrcrly snatch them up, feel of 
them with the lips or harhlct, and then hite into the <relatin. Thev imnuHiiatelv 
reject the <^elatin and they never repeat the prcn^ess. Kvcn if they dniw the tins or 
Iwirhlets rei>eatedly over the shells and the contained jrelatin, they never ajrtiin pay 
anv attention to th(Mn. 

I also re|M*ated with the hake the exin^riments which 1 had previously carried 
out uix>n the cat-tish, usinjx Ji tine-jK>inted pi|H»tte and s:ipid solutions. The tishe.s 
were in all cases rirst teste<l with sea water taken from tht» tank in which thi»v were 
swimminjr. On one o<*casion (the tirst test made) a jet of water directed ajfainst the 
filamentous dorsal was followed hy the chanicteristic Iwickward movement of the 
tish, so that he Anally n»ceived the jet in the face. He turned and tried to bike the 
]X)int of the pi|M'tte in his mouth -a puivly tactile reflex api>arently. This ies|K)nse 
1 never <rot aifjiin with this or anv other tish, thou<rh (H*<*jisionallv the fish would 
stop, hesitate a moment, and then swim on, i):iyinj^ no further attention to the 
stimulus. If the jet of water is directed aj^iinst the [ndvic fin while it is extended 
and searchinjif the lM>ttom for food, the tin is usually quickly withdniwn and pre.ssed 
ajrainst the side of the IkkIv. 

The pi|>ette wjis th«Mi fille<l with the freshly pn^pired and strained jui<'e of the 
mus.sel (J//'///V/A/), and this was dire<-t4»d aj^ainst the fish in tlu» siime way. The fishes 
resjKmded instantly, just a> when stiniulat4>d by meat, whethiM* the jet wits directed 
a<r»ii"*^t the filamentous dorsjil, or the dorsal fin at anv ])art, or the si(h» of the IkxIv, 
or the free }><'lvic fin. The n^flex i>; inunediate anci unmi>takable. more shai*ply 
<lefin<^l than I usually <r<'t hy contart with the meat of the sjime nuissel. The experi- 
ment was many times re]K»ated, alway> with the n»sult that the jet of water was 
iofnored or avoided, while* the jet of nuisMd or clam or cnih juice was eajrerly sought, 
the fish usually snapping at the end of the pijH»tt4». 

1 have carritMl out no systematic* chemimi exji<^riments to determine the gustatory 
preferences of the fi>hes, having shajMnl my ex|xn-iments so far Jis jiossihle ah)ng the 
lines of the n<u'mal fe^Mling habits of the species studied. Nagel and some other 
previous students of these prohh»ms have relied chiefly on reactions to unph^i.<^nt 
stinndi, and the reader is refern»d to their works, though I consider this a less satis- 
faetorv line of in(iuirv than the studv of normal reactions to food substances. The 
few frsigmentarv observations which 1 have made with chemicid stimulants I shall, 
how^ever, record in their appropriate pla<*es. 

S|MM'imens of hake were tested with a m.*> per cent solution of hydi'ochloric acid 
made ui) in distilled water, the a<'id lMMn*r directed airainst the IkmIv bv means of a 
fine pi|>ette. The doisjd and tins, the sides of the lx>dy, and the lips were 
teste<l. When fii^st tested on the fins one hake turneil and tried to take the pi|iette, 
much its he did with the clam juice. Afterwards this fish, jis well as all the others 
from the first, seemed rather to dislike the acid and would swim slowlv awav. There 


is no constant reaction, however, and in fact the fishes act verv much as tliev do 
when a jet of simple sea water is directed ajifainst them. Th<\v do not appear to 
dislike the acid intensely. Later 1 tested these lishc»s with a 1 per cent solution of 
hydrochloric acid in sea water. This is decidedly unpleasant and is uniformly 

The (^xj^'riments recorded seem to show clearly that the hake receiv(»s both tiic- 
tile and mistatorv stinmli hy means of the free fin nivs and to some t^xtent doubtless 
by other parts of the outer bcxly surface. What role may be playiul by the sense of 
smell remains obscure. To test the ]M)wers of h)catin«r conceahnl focMl the following 
experiments w^ere tried: 

In a tank cont^iininjr two hake which wen* very hungry I jilaced a piece of fresh 
clam meat concealed betw(»en two small, old. and thorouijfhly cUmui clam shells which 
had been lyinjr for some time in tin* bottom of the tank. Th<» tish(»s did not seem to 
smell the im^at at a distance and so Im' attracted to the spot where the shells w(M*e, but 
if in the course of their aimless moyements alon^ the liottom of the tank they passed 
oyer the shells, they f^enerally stopp(»d a moment, smclled around, and then passed on, 
fii*st feeling oyer the whole area of the sludl with thcii- free fins. As time passed, 
this reaction be<'ame less clear until after some fifteen minutes they «^enerally passed 
over the shells without payinj^ any attention. They nev(»r found tin* meat. This 
experiment was many times repeated with the same result. FIk* sens(^ of snnOl can 
play no strong" part in the locating of their food. It may |)lay some small part, 
though I incline to believe that the interest which th(» ti>hes sliow in the c(mcealed 
bait is excit(»d by a vague stimulus to the terminal buds on the tins. Compan* the 
experiments made after (»xtir])ation of the olfactory organs in the tomcod <lescri)>ed 

Thr toiiK'oil (Mtrrn(jii(lm< fonuod). 'rh«»s<» fishes are nuich less activ«* than the 
hake, spending most of the time lying (juietly on tlie bottom of their tank. They 
have not so keen sight as the hake and ])ollock. but still obta'ii much of their food by 
this sense, catching focKl thrown in before it reaches the lM)ttom. They do not catch 
live prawns in captivity so well as the hake do, yet pniwns and other active crusta- 
ceans are found in the stomachs of specimens tak(Mi with the seine. Tin* dorssil tin 
lacks th(* fn'e filamentous rays and is not ('specially s<»nsitive to gustatory stinuili. 
The v(Mitnil fins are, however, verv etficient in locating: sai)id substances Iviuiron th(» 
bott^MU. They are shorter than those of the hake and are not thrust forward, but 
incline slightly backward. Like the hakt», the tomcods spend nuich time in slowly 
i»xpK)rinjif the bottom, though they assume a very di fie rent position, with the head 
direct<Hl dow^nward at an angle of some ,*>() to 4r> with the bottom, so that the tips 
of the barl)l(»t and ventral fins just dmcr the bottom. When food particles an» hn'ated 
they are snapped up by a (piick lat(»rjil movement similar to that of the cat-fislu»s. 
Sometimes, however, stinmlus of the ventral fins is followed by a revers(Hl swinuuinir 
movement, the fish backing up to take the ))ait. At other times the fish when explor- 
ingf th(» lH)ttom swims slowly backwanl. so that no change of direction is necessary 
when fo(Kl is located. 

I made a series of tests with cotton wool and <'otton di|)p(ul in clam juice similar to 
those described for the hake, and with the sann^ results. 1 also repeated the tests 
made with sea watiM- and with stmined clam juice by the aid of a pii)ette, with iden- 


tioally tht' sanir result^i as with the hake. After a few tests the fishes ignore sea 
water and plain cotton, hut invariably respond to cotton soaked in clam juice and to 
the juice itself as thoy do to meat. The tomccxl reacts to bits of clear g-elatin soaked 
up in water essentially as the hake do<^<. 

I also tested the tomccnl with hydroi'hloric acid, (».2 i>er cent in distilled water 
and 1 per cei»t in sea water. Both an* obviously avoided. I tilled a tine pipette 
with a solution of quinine sulphate in sea water, al)out 0.1 jx»r cent — a very bitter 
solution. The tomcod swims away immediately if applied either to the lips or to the 
I>elvic tins, but appears not to notice it if applied to other parts of the IhkIv. 

Within two old clam shells, which had In^en lyin^ in the tank with the tomcods 
for sevenil days and had remained unnoticed, was phu-ed a piece of fresh clam. 
They were then closed to«rethcr and laid on the l^ottom of the acpiarium containing a 
tomcod. Shortly the tish passcnl near it, appeared to perceive it, turned from his 
course, and passed and repassed the spot until the shell was located, api>arently by 
smell, by a method of "'trial and error." Then he rooted at the shell vigorously 
until the two halves were separated and he could get the meat. 1 rej>eated this with a 
piece of s4|uid w ithin the shtdls with the same result. 1 tried two emptv shells in the 
same way. lie siiw me put them into the water, came up to investigate, smelle<l (?) 
of the shells and went away without so much as touching them, and never came back 
to them again. 

These exp<»riments were rei)eated in many forms luany times. In most of these the etiicient orgjin in discovering the pres^Mice of th(* iood was almost cei*tainly 
the pelvic tin. At least, this alone located it, for the fish swam alH)ut (jx>ssibly 
feebly smelling something good), but did not make a definite movement toward the 
Iwiit until the tins were dragged over the ciiick Iw^tween the two shells containing it. 
from which the juiciv^ were doubtU'ss IxMng ditfuscMl out into the surrounding water. 
Then he backed up in the typical way. If the* Iniit was not found within a very few 
minutes it was left unnoticed. ev(Mi though subsecpiently uncovered. 

The>e fishes almost invariably find a concealed bait, though the hake mrely does 
so. The hake seems to perceive th(» inlor or siivoi* of th(» f(MKl, for he lingers about 
the s|)ot where it is concealed, but never makes a movement to unco^'er it. The 
tomcod. on the other hand, actively pushes things about with his snout until the bait 
is disi'ovenMl. Hut. mdike the gjidoid fishes which Bateson descrilK?s, these fishes do 
not ifet the scent af the food at anv considerable distance and then search for it. 
Thev do n(»t notice the bait until within a few centimeters of it. and there is no 
evidcMice that the st»nse of snudl assists at all in the localization. 

To t(vst this j)oint the olfactory organ was extii^pated in sevei*al tomcods which 
had given the reaction last dcvscrilx^i clearly. Several ways of performing this 
operation were tried. The most successful m«*thod was to etherize the tish suflSciently 
to k«»ep him (juiet and then oix'nite in a shallow tniy with the mouth kept under 
water, cutting off the olfactory nerves or crura with a shaip scalpel. The wound 
suppunite<l badly, but appeared to give the tish no serious trouble, as they feed 
noi mally from the second day onward. Without going into the details of the observa- 
tions, I may say that after the third or fourth <lay the fishes took their food in all 
resptH'ts like uninjured fishes, so far as could be observed. They gjive all of the 
characteristic reflexes that have Ijeen mentioned alM)ve, includintf the dis^'rimination 


botwoen cotton wck)1 and cotton (li|>i>od in cluni juice and hctwoon sea water and clam 
juice aj)j)lied with a pii)ette, etc. The operat(»d iish would loiiite a conceah»d \ni\t ])v 
means of the i>elvic fins exactly as the normal tish does, an<l he would similarly root it 
out and eat it. In short, the j^ustntory refl(»x(»s, so far as 1 haye ohseryc^l them, woi'e 
al»solutely unmoditied by the o|>(M*ation. That the olfactory a])i)anitus wjis totiill}' 
destro>'ed wasyeritied hy autopsy dissections made aft^M* the close of the o])seryationH. 


T/tr M,a-rf»^tht (Prionufiis i'iirnli)nif<), -Tin* thn^e fin<,'(»r-like rays of the j>ectoral 
fins of the j^urnards hayc* lonjif attracted the attention of zoolot^^ists, and the American 
species of Pr'tittntinx hayc heen made the su)>ject of a can^fid res(»arch hy Morrill 
("05). He finds that, as in the closely related Kuropean 7/ •////</, tin » free niys are 
totiiUy deyoid of terminal Imdsor other specialized sense orpins and that the sensory 
neryes with whi<'h these free niys an* so abundantly su])])li(»d end free, like tactile 
nerves in j^enend. 

He also mad<» some interestin<r physioloofical experiments. The normal fcxnl of 
these species, so far as known, is small fish, youn<r <'lams, shrimps, amphiiKnls and 
other small crustacea, squid, lamellihninch mollusks.uimelids, and s(»awee<ls. (Linton, 
11>01, p. 470.) They are constantly fcclinj^ about the sand, turninjr oy(»r stones and 
feelinir under them, etc., with these free rays, and undoubtedly find their f(M)d lai'wly 
in this way, especially the ann«dids, mollusks, and crustacea; but in captiyity the 
eyes an* used chiefly in seciH'inj; the focxl. Morrill writers furtlu»r: 

hi onl«»r to tt'y^t tlu' ii«* <jf the fri'<» niys iii<lt^iH*n<U»ntIy <jf sij^ht tlit' crystalline lens aii<l c*<)rnca 
wen» n'inovcMl inmi hoiih» lisli, ami in otlior <*jisc»h the citrni'a \v:u< iMivennl with variiish, Uilsani, or tar. 
Till' rc|)eatr<l ('X}K'rinM»nt.« wen^ n<»jrativr' in tlieir result, as the lis!) |»ai(l no attention to tl»e f<MMl, even 
when it was }»la<HHl in contact witli t!»e irct» rays. 

Morrill con<*ludes "•that the free rays haye In^en nuxlitied for tactile purjx)ses, 
and that they are msiinly, if not altowther, used in setirchinjf for food. ■' 

ft •' if ^ r^ 

MorrilTs dissections h^aye it imcertnin whether the fr<»(» niys of the |HM*t^»nil ims 
receiye conummis neryes, as then' should do, of coin'se. if these orcTins had iriyen 
eyidence of ernstnt^>ry powers. The only scmrce of conummis filKM's for this fin 
would be throufrh the ramus lat^^nilis accessorius (r. nMMH-rens faciialis). Stjuuiius 
(1H41K p. IJ>) did not iind this nerye in Trlfjlti (jumanhiH and T, hfrumln, I dissected 
a si>ecin]en of PrioHofwH vmuilhtux and found the same to be true hen», so that it can 
})e taken jus assured that no conummis neryes reach tin* iM'ctoral fin in this s|)ecies. 

After an examination of the feedine^ habits of the adidt sea robin and of younir 

•* ft ^^ 

specimens alnrnt 10 cm. lon<jf I «|uite ae^ree with Morrill that the nMution to fcKxl 
particles by the free fin mys is tactile oidy, with no gustatory elenuMit. When 
adults are fed with fresh clams or muss(ds, the shells split ojxmi to exjH)se the meat, 
they turn and bit(^ out the meat as soon as a free ray touches (he soft llesh. Vounjr 
fishes <lid not e^iye this reacticm so inyariably, and eyidently relied nuu'h more on 
sij(ht. C'lean clam shells filled with mcdted ji^elalin wen* reacted to like the fresh 
clams once or twice l)y (»ach fish, but usually were thereafter i<»'nore(l. 

The free niys constantly stir up the sand and jifniyel of the lM)ttom. If soft 
edible j)articles are t-ouched the head may b(» tiu-ned to snap them up, es{x»ciall\^ with 
old fi.shes. With younjjer ones this usually do(»s not happ<»n unless the particle is seen 

THE 4)R(;aN and sense OF TASTE IN FISHES. 265 

while in motion. In fact, with these youn^^er lishes the purpose of the activity of the 
free rays seems to l>e in tlie main the ag^itiitiori of particles on the lK)ttom to bring 
them into the nin<re of vision. Ahnost any unfamiliar object, such as a bit of coal or 
a brijrhtly colored j)e])ble or any soft i>article, if s(»en while in motion, will ]>e apt to 
Im» taken into the mouth. The analysis is done here — not by the ixMiphenil cuUmeous 
orj^ms. All small objects thrown into tlu* water are taken into the mouth as the}' 
fall; bits of filter pa|)er, <relatin. etc., will be taken and immediately rej(»cted. The 
sjime bit of paj)er or excrement may be taken and rejcvted a half dozen times in 
ra])id succession, the reflex followin<r in a jKMfectly automatic way as soon a.s the 
movin*^ ol>ject is secMi. Small worms when thrown into the wat4*r would l»e captured 
lK»fore thev had time t<^) rearh the bottom. I)ut if ])la<-ed on the liottom thev would 
seek shelter under pebbles and remain uiuioticed luitil they were stirred up and sent 
Hojitin<r otf. when thev would be seen and taken at once. The free* lin r.iv wjis 
oljserved to touch the worm when concealed without evoking a response. A moment 
later the worm was set in mf)tion and taken at once. 

1 got no evidence* that the tish(\s smell or otherwise*, (h^tect the presence of food 
at a distance* or concealed from sight and touch. Meat inclosed lietween <*lam shells, 
which a tomcod would have secured within a minute or two, remained unnoticed, 
though the outsiders of the shells were rep(»atedly fingered over by the free rays and 
similar bits of meat were taken at once* if in motion ne^ar the fish. 

The vountr sea-robins eat crab me»at wedl. 1 mtide* a stronif e^xtract of crab me^it 
and tiltereMl it. Now with a fine pipette a jet <jf e-le»an seni water was elirected Jig:iinst 
the free pectoral -fin rays. TheM*e' was no re*sjM)nse, or if the jet wa> strong the fin 
was folde»d agiiinst the' l^ody. The e»xtraet of crai» applied in the same* way with the 
pil)ette gJive the sjime* le^sult. KveMi whe'ii tbe» jet is direH*te»d against the lips the fish 
usuallv ])avs no atte^ntion av is distuibe'd anel swims awav. This would seem to 
indie*:ite* that the' se'use* of taste' i> al)se'nt or ve'rv feM'})le* on all of the* exiH)seMl jKirt.s of 
the ImhIv. Thus the' absence of spe'cial gustatory se'use organs, e^f conununis nerves, 
and of irustatorv reaction> fiom the* fre*e r.ivs of the* pectoral fins serve as nmtual 

^ • * I 


77/' hiiifj-psh (Mrnfirirrhn.s sti.rtif IJ 1.^). The*se* lishe\< have a sliort, thick mentiil 
Iwirblet, and the*y were studied to e'fjinpare* the*ir re*actie)ns with those e)f the siluroid 
and gadoiel lishes. Most of the* tyi)es of experiment made* previe)usly on the* latter 
tishe\s were* re|M*ateel e>n the* king-fish. Without going inte) details, the ex|>4*riinents 
seemed to show in general that the king-fish is not a pure visualize*r, though vision 
is somewhat used in fineling food. This seems to be in the main a tactile reaction, 
as most of the fexnl take*n was by e-ontjict and nonnutritious substances were 
general I V taken if thev fe*lt like food. For instane-e, colorless gelatin is taken at the 
first ce)ntact and re|x*atedly thereafter for an indefinite muuber of times, though \n 
each cjf^e it is at once rejee'ted as s(K)n as it enters the moutti. The sense r)f taste 
seems to l)e liniited te) the mouth, and I found no evide»nce of a gustatorv reaction ))v 
the l«rblet, though the ex[>eriments we*re not sufticiently numerous or varied to l)c 
conclusive. They do not find a concealed bait. 

77/r t^nnl-iish {(^p**m}*i>< //7?/).— These fishes were exi)erimented upon at the .same 
time as the hake and tomcod, and bv the siime methods. The toaelfish never found a 
concealed l>ait and never seemed to get food by any other reflex path than the visual 


or tactile. The fleshy, eututieoiis appendajifes of the skin were especially U»st4^d to 
hring out possi])le gustiitory reactions, but with negative n^sults save for those* bor- 
dering^ on the lips, where it was impossible to exclude the participation of tast(^ buds 
on th(» lips. This agrees with the anatomical findings of Miss Clapp (181)i>), whose 
careful studv of thc^ skin of this fish faih^d to reveal anv terminal buds on these 
appendages or elsewhere away from th(» buccal cavity. A j(»t of sea water directed 
against these appimdages or the IkhIv -iurface in g(»neral usually disturbs or frightens 
the animal meri^ly, if it is noticed at all. A jet of clam juice similarly applied i»alls 
for the same reaction unless it is so directed sis to n»ach the lips, in which <*ase the fish 
reacts to it just as the hake and tomcod do, attempting to take the tip of th<» pipette in 
the mouth. The following solutions were applied in the sanu» way by a fine pipette 
to various pirtsof the body surface: O/J per cent hydrochloric and 1 ]MMcent hydro- 
chloric acid in sea wat4»r, and 0. 1 p(»r cent (juinine sulphat(» in sea water. In all cases 
the fishes paid no attention to the stinuilus unless the substance was so applied as to 
come into contiict with the lips. The exjxM'iments lead me to conclude that the toad- 
fish can taste only within th(» mouth and on the lips, and that if th(» cutaneous api^n- 
dagres have anv sensorv function it is tactile onlv. 


The morphological and physiological significance* of the tenninal })uds of fishes 
is a pro})lem whicli has exercised some of th(* ai)lest morphologists for over half a 
century. The methods of the older anatomy have signally failed to yield concordant 
results. Not until the innervation of the cutaneous sense organs was worked out 
from the standpoint of nerve components was this confusion relieved. The older 
morphologists (Schulze, Merkel, and others) discovered a morphological criterion, 
the '*hair c(*lls," by which the terminal buds could b<* distinguished from cutaneous 
sense organs belonging to the lati»ral-line system. But this fact attiiined its signifi- 
can(*e only when it was discovered that the organs of the lat(M*jil-line system, or neu- 
romasts, which possess the *' hair cells,'' are always innervated by lat(»ralis nerves 
related centniUv to the tub(M*culum acusticum, while terminal buds, which lack the 
"•hair cells/' are always innervat(»d by comnnmis nerves which are related centmlly 
to the primary gustatory centers of the vagal and facial loi)es. 

Presunuibly, then, lateml -line orgjins and terminal buds have ditferent functions; 
and, furth(M*, the function is probably not tactile in either <".ise. sinc<» all parts of the 
skin receive general <*utaneous nerves in addition to the sp^M-iai sensorv <*omponents. 
and these general cut4ineous nerves are relat(*d proximal ly to ditTf^nMit ctMiters from 
either of the oth(»rs. The lat(»ral-line organs are known to be used in the maintenan<*e 
of }K)dily e(]uilibrium and the perception of mass motion of the water. (Compare 
the recent works of Iah' and Parker.) On the other hand, the terminal ))uds are 
related in structure and innervation to undoubted taste buds of the mouth, and hence 
the inferen(re that their function is taste. This inference is abundantiv confirmed })v 
the experiments h(»re recorded, and the function and morphological nmk of the 
terminal buds are at last definit(*lv fixed. 

It may be regarded as established that fishes which possess tenninal buds in the 
outer skin tast^* by ujeans of these organs and habitually find their fo(Hl })y their 

THK or<;an and sense of taste in fishes. 2^7 

nu'ans, whilo tishcs which hick those orjipans in the skin liavo the sense of taste 
confined to the mouth. The delicacv of the sense of ta^te in the skin is direetiv 
proportional to the niunher of terminal buds in the areas in (juestion. 

Numerous unrelated t\\ws of }>ony fishes from the sihiroids to the ^doids which 
jK)ssess terminal huds hav(» deveio^xnl s|)ecially modified orcfans to carry the buds and 
increase their efficiency. These orcfans mav take the form of inirblets or of free 
filiform fin rays. The free i*ays of the j^'ivic and dorsal fins of <r«»doid fishes are thus 
explained, and indeed this is possibly the motive for the mi^nition into the jugular 
|)Osition of the jxdvic fins of the t^adoids. 

In all cases where terminal buds are found on })arblets or filiform fin rays g-usta- 
tory nerves U'louj^in^ to the communis system are distributed to them. These 
F)arblets and free i\n niys likewise receive a very rich innervation of tactile or gen- 
eral cutaneous nerves, so that they merit their popular desii^nation" feelers." 
lioth sets of end organs undoubtedly cooixrate in the discrimination of food, and the 
animal has the power of very accunite localization of the stinudus. Whether the 
jifustatory stimulus alone can 1m» localized apart from its tactih^ ac<-om])animent can 
not at present be .stated. A purely tactile stimulus with no <rii^tatorv element «in 
be localized precisely, and 1 have as y(^t no conclusive evidence that a pure gustatory 
stinudus. even when strong, can l>e located by the fish. It is certain that feeble and 
widely diffused t^ustatorv stinmli <*an not be accunitelv located bv the fishes which 1 
have experimented with, either by the terminal buds or by any other orgsins. 

The fishes in which th(» cutaneous tenuinal buds are most highly develo{)ed ai*e 
in general bottom f(*edei*s of rather sluggish habit, and in some cases they are noc- 
turnal f«MHlers. The high development of this scum* is compensatinl for in some 
fishes by the reduction of otluM's. The visual i)ower of the fishes is especially apt to 
suffer d(»gnidation. Thi> degradation may U' orgimic. a iM)sitive degeneration of the 
visual apparatus, a- in .1///' ////'/n. or it may Jh^ merely functional. In the latter 
citse, thouirh the organs of vision are not necessjirilv modified, these organs are not 
actually used in prcMuring food, the fish InMug unable to eff(»ct visual reflexes toward 
f(KHl substances or to correlate visual stiundi with th<* movenjent> necessary to react 
toward food substances. Tht» fi>h mav be ])erfectlv able to eff(»ct oth(M* visual 
reflex(*s. })ut is apparently unable to understan<l the significance of foo<l when per- 
ceived by the sense of sight only. This ]iarticular central refl(»x i^th has never })een 
deyeloi)e<l, or ha^ atropbi(Hl from disuse. Nature ha** here eff(»cted for the s|)ecie.s 
something similar to what is accoujplished in indivi<iual men oc(*jisionally by <liseas<\ 
in the pnHluction of certain apha>ias. 

The nuudw^r of reflex activities habitual to an animal with a nervous system as 
simply organized jis the bony fish is probably far smaller than is commonly supjwsed, 
and these activities are in general <'haracterized hy but little complexity of orgsmi/ji- 
tion. It is prolwibly cpiite within the nnig(* of |K)ssibility to determine by observa- 
tion and ex|)eriment for any given species of fish, to a high degree of accuracy, what 
these halntual activities are and to work out bv histological nu^thods the reflex arc 
within the nervous system for each of then): and since the human nervous .system is 

ft- * 

built up on the sjime general plan as the piscine nervous sy st<Mn it follows that such 
a thorough and systematic correlation of function with structure would Im* profitable 
from many points of view. 

. I 

? I 



' I 





Duriiii;; tho winttM* and sprinjr of iwy, some furth(?r <)l)Si»rviition> have lieeii 
iimde with th(» puriK)si» of iiiisweriii^ (umon^ others) the ((uestioii raised almve, 
wliother fishes can h)ealize a sensation received l)v the terminal huds alone with 


no tactile aeeonipaninHMit; or, in other words, whether j^ustiitorv sensations may ]>e 
provided with a local siirn as tactile sensations are. (This (juestion, of conrst*, does 
not necessarilv involve the more ireneral one as to the essi^ntial nature of the loi-al 
sign, whether it is due to a "'simmmHc energy" of the periphend nerve or s«»nse organ 
or to central differentiation in the terminal micleus.) 

Some rt»cent clinical observations suggest that in human lM»ings such a l<H*aliza- 
tion of gustatory sensations is possible. Crushing (Johns Hopkins Hospital Bulletin, 
vol. XIV, No. 144, 1JM»8, p. 77) n^ports after destruction of the (iass(»rian gsmglion 
and totjil pandysis of general sensation on the anterior part of the tongue, that the 
gustiitory sensibility remains unimpaired, and that in this case the gustatory sensji- 
tions can be localized. It is not. however, absolutely certain that it is the gustatory 
filx»rs which effect the localization, for the chorda tym[>ani. which was uninjured, 
may carry also a certain nundier of fibers for gent^ral s(Misation from the facialis r(M)t 
in addition to gustjitorv fibers, as Chishinj^ assumes is the cas(» with the chorda from 
some of his results anil from those of Kftster. 

My own observations were made on the young of Atnrliu'tix from ."» to s cm. long, 
received from the Stale fish haU'hery, at London, Ohio, in OctolnM', 11H>2, and kei)t 
under obM»rvation in tanks during the following winter. These fishes prove to be 
more shy and less teachable than the smaller .b//<////'//x fry (alKKit 3 cm. long) hatched 
}>y wild i>an»nts, upon which the exiH»riments rc»|)orted in the precinling pages weiv 

I have verified on these fishes most of the obst^rvations made on the smaller 
fishes last vear. The most notictMibh* differt»nce in their lM»havior is the evidentlv 
greater visual power in thes(» fishes. As soon as they begjin to feed freely in the 
presences of the o)>server (which nMiuinnl several months of training) they l)egan to 
show evid(»ncc of visual recognition of a moving bait, if V(»rv near them, and pro- 
vided they had just previously l)een fed with the same focnl in the same way. Thi»v 
never undt»r anv circumstances notice visuallv a still bait, and tlu^ir recognition of a 
moving bait is at best vei'v imperfect and only an occasional (H*curren<'e. 

l'lM>n putting a concealed bait in a tank with tlu» fishes I found no evidence that 
they are able to locate it by tin* sense of smell or otherwise from a distance, pnivided 
the water is still. If, however, they swim nejir enough to the ca])sul(» containing the 
bait (lK»ef liver, cheese, i^tc.) to pass the Inirblets into the strong diffusion currents 
emanating din^ctly from the bait, it is h)cated instantly. The reactions h<»re are 
essentiallv like those bv which the tomco<l localizes a concealed bait, though I have 
not com[)leted the experiment by extirpation of the nos<^ to determine what part, if 
any, is played by th<^ sense of sm(?ll. So far as my exiK^riments have gone these 
fishes will not locate a concealed bait in still water unless they i)ass within 5 cm. of it. 

In ruiming water, howev(»r, the case is quitt^ different. I constructed a long, 
narrow tank, so arrang<*d that a slow stream of water can pass through it from end 


to end. By roverin^ the lower end of the tank and iiluuiinating moderately the 
upper end. it can be so arnintred that tht» nepitive photot.ixis will countei'act any 
positive rheotaxis and the rishes will remain in tht» lower end of the tank. If now 
liver or other stron^r l>ait is plaml alK)vi' them, the tishes will promptly swun up 
the current and loejite the meat. 

The exiH'riments >eem to indicate that euncealed footl can not he located by thcM* 
tishe.s from a distance in quiet water (cf. Natrd. l.y.*4). hut that if the fish passe*^ 
within a few (•entimeter> of it the ditTu>ed juices are recot^nized and the focnMocated 
promptly. In ruiminj^ water, however, the rishes will follow the diffused juices up 
the stream for considenihle distances and so rind the food -a fact well known to 
everv fisherman. Tactile >eiisation> are elearlv not involved: it lies l>etween the 

ft « 

senses of smell and ta>te. and I have not as vet i^one far enouirh with this series of 
exiK»runents to decide finally the ynwi played hy the sense of simdi. 

I have, however. te>ted the sensitiveness of the harhlets to diffused savors 
moi'e fullv. Raw meat or beef liver was minced, extracted in a little water, and 


strained. A wisp of cotton was wound on the end of a sh^ider wire, dippi^d in the 
meat juice, and gently lowered so as to lie a few milliujeters from the tip of a harhlet 
of a cat-fish which was otherwi>e entirely concealed under a large leaf. The tish was 
upahle to s«m» the cotton and actual contact with the harhlet was carefully avoided. 
Within a few seconds the fish hci-ame conscious of the savor and turniHl tnimrd th*' 
Ofttoii. Again. 1 filled a glass tuht*. of aiK)ut *>-nnn. ijore. with the meat juice, closed 
the upper end with the finger, and carefully lowereil the open end down over a pro- 
jecting harhlet, as in the previou> ca>e. Tlu' specific gravity of the meat juice is 
slightly greater than that of the water, and from the lower end of the tulx^ (the upper 
end l)eing ke[)t <'Iosih1) the jui<e *»lowly diffused downward enveh>ping the tip of the 
Imrhlet. without, however, any noticeal»le current IhMng produied in the water. The 
fish locates the stimulus and turns toward the source of it. In other cases I colored 
the juice with a little hloixl. >o that the course of th«* (liffu>ion curn^its could W 
o})served, and it is evident that the reaction follows the >timulus of the hnM^t only, 
and not the organ of smell, for the ujovement is made lK?fore the diffusion currents 
have had time to reach the nostril. 

These reactions are not as proujpt or precise as those given after a cnntuct with 
a siipid substance where a tactile sen>ation accompanies the gustatory, and in a large 
percentage of the caM'> then' i> no definite reaction toward thi» point stiuudated, but 
merelv the more vague ••seekiuif reaction" to which reference has l)een madt* above. 
Nevertheless they indicatt* on the whoh» that pure gustatory stinudi, if very strong 
and applied to a small ai'ea of the |x*rcipient organ, rnn h* 1*Hui1iztd !n ^jntc* . f*r h^irt' // 

JLn/ JO, JfJOJ, 



Allik, E. p., jr. 

M)7. Till' cranial iiiUiM.4('i4 ami rmiiiul and lirnt Hpinal ni*rvi« in Ainhi culva. Jonrn. Mtir/th., xii, .'{. 


'00. Thti wnsiMnxans and iH'nx'ptionn of lislu's, witli n'markn on iUv. snpply of Iwiit. Junrh. 

}ftinnr liiol. Jxvof., Jjiifuhtu, I, pp. li25-256. 
'{HUt. S<»nnt» of toiK'li in tlii^ nH:klin{;. ////(/., p. 214. 
Clatp, Counelia M. 

*JilK The lateral lint* syt<tvni of Unhnrhun tan. Jonrn. Mtn-fth.^ xv, 2. 

(iUADKU, V. 

'85. Ver>rlei(:hen<U' (irundversnclie iilKjr die Wirknnjj: und <lie Aufnalini.*<t('llen cheniisclier 
Iveize Ihm den Tien^n. Jilnl. (\nt., VA. v, Noh. VW, 15, MS. 

*8i). llelxT die Kinptindlielikeit einij^T MiiTtien* j^'j^en Kieeh«toffe. //m/., I»«l. viii, ]»p. 74.S-754. 
(itTNTHKK, A. C L. (J. *H0. An introdnetiun to thesUidy of iishen. J'^'iinhtnt/h. 

IIakkis, Wm. C. H)2. Siilnum ami trout .Vinerie^n Si)orl,^nian'K Library. \. )'., Tin Mamilftnn f'o. 
IIkkkk'k, C'. Jri).»«)N. 

'91^ TlHM'ranial and iirHt spinal nervi^ of J/ir/iiV/Za; a contribution u{M)n the ia*rve ('otn|x»nentH of 
the lH>ny li8h<*H. Jonrn. of ConifHir. \rnrof(H;if^ vol. 9, pp. 15.'{-455. 

'(M). A contribution uiM»n the cranial nerves of the c(Kl-(i8h. //m/., x, ',\. 

*lM. Cranial nervew and cutaiUMUiH senae or^rans of North Aniericiin Siluroid fi.^heH. Ihiti., \i. 


'72. Etu<lej4 <ranatoinie coiniMin'e Hur h*s orjranen du toucher chez diven* tnannnifi-res, oiseaux, 
|H)iHHonH et inscH'tes. Ann. sr. mit., 5 S4*r., T. xvi. 
Kaiilknhkiki, L. 

'98. The action of solutions on tlu^ kmis«» of tast<;. />»//. I'lur. of \Visritnj<in, S<'h nir Si ri,s, vol. ii, 

1, pp. i-:n. 

Lek, Fkeokkic S. • 

'92. IJelHjr <len <Jleichj;ewichtssinn. Crntrafhi. f. /V///x/o/.. lid. (», pp. 5<)8-5l2. 
'9,*{. A study of the sensij of (MjuiJibriuni in lishe.**. jnir. of Plufsiol.^ vol. 15, pp. ;{lI-:»4;». 
'94. A study of the sense <»f iHiuilibriuni in lishes, l*art ii. Jonr. of I'hifsiol., vol. 17, pp. 192-210. 
*98. Functions of the ear and lateral line in lishes. A nor. Jimr. of rhj/sutf.^ i, pp. I2.S 144. 
Lkviiic, Fk. 

'51. IJelKT die au.«Hc?re llaut oinijjer Si'isswasserlische. Z*'ifit. y/vVw. /f;o/., m. 

*79. Nttue Beitnige zur anatomis<'hen Kenntnis.H der l!autde<'ke und Ilautsinnesor^ane «ler Fi.^che. 

I'VuUvhr. z. JfH) Ji'tkr. Nntnrf. (ies. :n IlnUr. 
*94, Inteiirunient und llautwinnesorjrane der KmH-henfisc^he. Wi'itere IJeilrajn*. Ztntl. Jhr. .{!»(. f. 
An4tt. n. (httoffrn.y viii, 1, pp. 1-152. 
Linton, F. *0I. Parasites of lishe« of the Woods I IoU' rejjrion. lUUl. V. S. Fmh Com. Un- Lsun. 
Mkkkkk, Fk. *80. UeU^rdie Kndijrunjjen dii* s<*nsiblen Nerveii in tier llaut der WirlH'lthien*. JiimttH'k'. 

Contiiins extensive biblio^rniphies. 
MoUKiLL, A. D. '95. Pect(jral apiKjndajres of /V'Wo/'/^ and their iiuiervation. Jonrn. MorplnkitMjif, \i. 
Nagkl, W. a. 

MM. Verjcleichend ^diysioh^jjiscln* und anatoniis<*lu^ rntersuchunjren uIkt den <Jenichs- und 
<TeHi^hniack8Sinn und i lire- Orpine, niit einleit^'uden Ii<*tnichtungen aus der all}^4>iiieinen 
venrleichenden Sinnespliysiohvit-' IliftHothvfa ZiMdoijuti, Sfutifftni, lleft IS. <\»ntains a 
biblio;rraphy of IiiJ5 titles. 

]*AHKKlt, (i. IL 

'0.'{.< of hearing; in lishes. Al»stra<'t of a i»aiH'r read U'fore the .Vin. .\ss<h\ fi>r the 

Advam^enient of Scienci*. St'h'nn\ N. S., vol. xvii, No. 424. 
'(KWi. The wnsii <>f hearinjr in lislujs. Ant. XiUnrnflM^ xxxvii. 
'03/>. llearin*; and alliiMl siMises in lishes. lin/f. V. S. Finh (.^nninlssioit for Jitfts. 


'98. The relation of the tast*i <A aci<ls to their de^rri'e of dissociati<iii. .1//*. i'linnnal Jonrwif. 
S<"iiri.ZE, F. K. 

'6.*i. Ut'ber die lKK»herforniijren Or^ane der Fische. Ztutn. f. n-'iHu. ZiH,i.^ xii, 2. 

'07. Fpithel- und Driisenzellen. Arrh. f. niikr. Annf., in. 

'70. UelM.T<lie Sinnesory:ane<ler StMtvnlinie Ikm Fischen und AniphibitMi. .\nh. f. mUc. Anot., vi. 
Stanniuh, H. '49. Das iwripheriscln' Nervensysleni <ler Fist-he, anatoniiH;h inid physiolojrisch 

untersucht. Ilimt^H-k. 
ToiJAKo, F. *7I{. Ia« orpines du jroiit et la niu<|Ueus4' buci-o-branchiale ties S<'*laciens. Arrh. Ztntl. 

Krjkrini.f ii, pp. 5.*>4-558. 
VVkiier. '27. Uelxir das < Jes<'Iiniiu'ksorpin der Karpfen. .yfrrkiTs Arrhir f. Atnit. 
WiiiTM.\N, O. (). **M). Animal iK'havior. Diolop<'al K^tures, WimmIs Hole, si*ssion of 1S9S. Iloalon. 
ZiNcoNE, A. '78. (VHservazioui aJiatouiiche su di aUunu^ apiiendici tattili ilei iK'sci. Itmtl. Amul. 

yupoliy XV. 




My H. *S. JKXXINCitS, 

.Insist, nit /*tttfts.\ttr i>/ Xtmiii'^y in fhr f 'ui:i't\<ify f/ Miihiirafi. 

F. '". II. r.nrj— iH ^»7:» 




»> II. S. JHNNIN<;S. 


Thei*e is i>frhaps iu> ii«*e<l so ^rivai in Aiurriraii zoology as to have the different 
jjrroups of iiivertel)rat«*s thoroiiixhly descrilM'*! ami s«*t in i»nhM\ so that the worker in 
ecolojry, physiolojry. varialitni.or ni<»rphoh><ry <*aii determine Iheni without l^ectmuii^ 
a pn)f«»ssional syst«'niatisl. As matters stand at the pivsiMit time, most of our 
aquatie invertebrates ean nut }>e determiiUHl without the study of niueh si-attered 
litemtuiv. aneient and modern, and mueh wearisome and unprotitable sifting of 

Happily thtnv is at pr^si'm a stronir movem«Mit towanl remedying: this stale of 
affaii*s. What is needecl is a s<'t of studies e«»mprising monograidiie treatments 
of the various ,iri*oup.s — eaeh ae<'Ount td« a ^roup eomph'tt* in itself, so far as the 
Amerieau si><mm«»s areetmeerned. so that any spe<-ies of the ;jrroupcan Ikj determiiietl 
without reference to other literature. This ean be aeeonipli.shed if diffeiXMit inves- 
tijTiitoi's seleet eireum.seril>ed •ri'ou[»s of not too ;rreat extent, perhaps a sinjrle jrenus, 
and s«M this thorouirhly in onl<»r. deseribinjr an<l fi^rurin^r all s[M.»eies likely to <M*cur 
in Ameriea, antl l>rin.ti;in;; thr names into etuisonanee with iveojrnized rules of 
iiomeiK'tatun'. It is suehasludyof on»» of tin? families of tin* ]{o1atoria that is 
herewith present^'d. 

The Rnfinllda an* a family of five-swinimin,^ Uotatoria, eontainin;: altogether 
aiNUit 4<» to 4."» s[MM'i<'s. Their chief general interest lies in their iK»euliar uiisym- 
metrieal strueture, most t»f them having the organs sodistH>sed as togive the impn*s- 
sion that the ImmIv has l)een twisted, while the primitive bilateral symmetry is still 
further disturlMMl by a numlMT of the organs lNH*oming rudimentary on one side. 
They are found as a rule amid a(tuati<* plants in the <|uiet parts of lakes, jwiids, 
and stivams. Only one <»f them {Ktiftvlus f-npurinus Wierz. it Zaeh.) can Iw saiid 
to lie limnetie — that is, eomnioiily f«uind five-swimming at a distanee from the vege- 
tation of the slumps and bottoms. A few oi'cur in swamps; but clear water, amid 
actively growing vegetation, is the place where the RniliilUUt al>ound. In such 
ivgious they aiv ofteu among the most abundant of the Kotifera. 



Tlu* olas^iliraUtui of thi» Ratlnlifhi has lalltMi into j^^n^at <*oJifiisi<ni. Tlii?> Malc- 
iiioiit roiiKl hr iiiatU'ttf almost any of tlu' larp'i'^roii|)s of Kotatoria. but it is|»erhaps 
nu»n» Mrikiii;;lv triio o( tliis fainilv tliaii t»f aiiv oth«»r. Maiiv siKM*ie> haw Ihmmi 
ih'sriilHMl iiihIit soxoral tlilTmMit s|M'rilii' ami iroiiorir iiaiitt'S, whiU* in otlirr inhm'S 
M'voral ililTon»nt >pfri«'> ha\o Im'^mi <U»srrilK»«l undtM- a .sin«rh» nanu*. Thi* t\viMe<l, 
unsymiuctrii'ai siriu'tun' has always l)o<»n moro ov h»ss <»f a puzzle t«i Mstrnuitists, 
luakinirit iliftifult w» ih'tt'rniintM»vcn what were proiMTly tt)]H» r<»nsi<h»n'«l doi-siil and 
vonlnil surfart's, and tho ^ifat ditTt»nMUM» in appiv'irancr ImMw^cu «'ontrarttil ami 
oxtondiMl animals has turtluM- trndiMl to favor ronfnsi(»n. It has S4M»imMl in tin* writer 
that thort> is no irnuip of tho liolatoria so mu«*h in mnnl of a tlnn-ou^rh n-vision as 
this ono. I'\»r this ivasoii it has Ikvu takm up lii-st. 

In thf t't>llowing pa]>«M' I attomj)! to givo an aoronnt «»f tin* Mruotuiv and inow- 
nn'nt> of thi»M» aninnd>« jKiyin^osjHM'ial atti»nti<»n to the asyinniotry ami it> hiolo^iral 
siirnitiranrr, anti to t*urnish as far as possihh* full drs^'riptitJUN and !iirur»'> «if all 
known siHH'ios, A largo nmjority of tho known s|mhm«'s I havt' mvM'lf }m^'1\ ablf lu 
study, and in tln^sr i-asos tin* descriptions ami figures are Iwised «in my own oh.MTva- 
tions. I have attempted to nuik<* tln*S4' so detailtMl that furthrr niiMakr^ in tin* 
idem itieat ion t>f thest* speeies will hanlly Ih» ]>ossihh'. In tin* east* tif s|>«H.-ieN whieli 
I havr not lH»en able Xo examine myself I give the tignri»s and «li'>rrii»tit>ns whieh 
have l>een ]uiblished by (M her authors. Many of tlnvM* des4'ription> an* very unsjitis- 
faetory, as comparison with a largtMiumlKM- of spt»eies i^ ur**i»**Hary for bringing out 
thr im|HM'tant charatMeristies. ami sneh ('omparistui ha^. in thf ab^^rui'i* uf pn'^M^rxfil 
material. U^en almost whollv laekiiii<: until V(M-\ refenl linir*^. 

In the preparation of this paper I ha\i' Ihh'u es|HM*iall\ imlebif^l lor a>si>iait«-«* 
*tf the most essential ehararter \o Mr. rharh».s V. Kousselet, of l.itmltin. Kngland.aiul 
to Mr. I'\ \{. l>i\on-Nuttall. of Keeleston Park, North Treseot, Kmrland. Mr. ]{oiis- 
selel plaeed at mytlisjH)sal hisvabuible nnuinted eolleetion of llu* A'''^'*./«f/ii . ineliid- 
ing a number of speeies whieh I ditl nt»t have in m\ eolleetion. and has as>isT«-«l iiu- 
thnuighout the work with valualde notes and suggest itMts. Mr. l>i\on-Nutiall M»nt 
nn» his notes anti ilrawings(»f a eonsiderable number o{' spt»oiesof Ihii-*-*", whii-h lit» 
ha<l long l>eeii studying, auil ga\e me permission t<» make um* nf si»me of hi*« ev^-^'l- 
lent tigui'es, a numlMM- of whieh are gi\en ou [dates t\ and \lli. Tin- finiinu^Ml 
e(H)|HM'ation of these twi» earefnl investigators has added mueh to tin* •-•»mpli>!**iit»s.<^ 
and aeeuraey of this paper, anil has made it |H»ssibU\ b\ oMuparison **l sjH-i-iniens, 
tolH»eertain that my dett*rminatH»ns of doubtful s|H*eie«i airi^'** with iln'se i»f th** Ih^jsI 
KumiKMU anthoriti<»s. 

I am indebted also tor sp«M'ini(*ns of /pr/Z/^Z/.i'ii from Lake ImiIo^zih* in liussia tn 
the kiutlness td' l>r. Homuald Minkiewie/., of the rnivei>ity of K;is:in, Uussia: 
to llerr Max Ntdgt, o! IMon, I am umb»r td>li^ation^ tor s|HH-imenN ..f h> lu-w 
s|H»eies, lUnrr/hi nmssthh. l«'or noies ;ind other assisiauiv I am indebu-i :o Hcrr 
OlH»rfoi*ster 1.. Ililllnger. of Slullgart. iiermanx : to V\>^i, P*. OVu* /aeiiarirts, 
din»et*n* of tin* l''r«\sli>\a!t»r llitdogieal Siatitm at Ph»n. itermany. a?n! t.» P:^•^ I>r, 
Kai'l Kekstein, Kberswalde. <«ernmn\ . It is a plea«*U!"s' :o e\p:\*s> heiv m\ '.banks 
lo theM' gentlemen. 




The use of preserve<i in<)unle<l spi^eiinens has been tJie basis for the present work. 
It is only through the methods devised witliin the last deeade by Mr. Charles F. 
Rousselet, of London, Englantl, that the use of such preserved material has l)eeome 
possilde in the study of the Kotatoria. Ilenee the complete lack hitherto of type or 
reference siM*cimens among tliese animals. This has Ix^en one of the prime causes 
of the great confusion in the classification of the Rotatoria. A few of the genera 
liave been worked over in the last fcwyeai-s with the use of preserved specimens by 
Mr. RiMisselet and his collaborators in Englan<l. It is not too niu(*h to .say that it 
will ])e necessar\' to go over the entire group of Rotatoria in the .same manner l>efore 
onler can \ye ])rought out of the present confusion. 

KiUing a:ul pre>if rmtion. — The collections of preserved material on which the 
following paper is based were made as follows: The Rotifera were taken in various 
ways — by towing with the tow n<»t in water free from vegetation, by washing aquatic 
plants in jars of water, by bringing into the laboratory ([uantitii^s of aquatic plants 
together with some of the water alwut them, etc. Most of the Rotatoria come after 
a time to the lighle<l side of the vessel in which the material collected is placed. 
These an* transferred in large numbei's to a watch glass an<l placed l>eneath a 
simple microscope or low power of the compound mi(*rosco|>e. where the movements 
of the organisms can Ix* obs(M-ved. 

Now a considerable <[uantity of Rousselet's nai*cotizing fluid is mixeil with the 
water in the watch glass. One-fotirth as much narcotizing fluid as tlien* is water, 
or a larger or smaller proportion, may be used, as se4»ms d<»sirable from observation 
of the movements of the animals. Rousselet's narcotizing fluid consists of '1 percent 
solution of hydnn^hlorate of c(H*aine. 3 i)arts: methyl alcohol, 1 part: wat^M-, fJ parts. 
This causes th<» animals to swim slowly and gradually to settle to the liottom; the}' 
will sixm die, and if allowed to die unfixHl will be ([tiite woi-thless forstudy, <le.struc- 
tivechanges taking plaee in the tissuesat the moment of death or i)erhapseven before. 
As .soon, therefore, as most of the rotifers have sunk to the bottom, as much of the 
wat4»r as possible is drawn otT from alM)ve them with a pipette. Then a small amount 
of n.2.") ]H'r cent <»smic acid is intro<luced, which kills and fixes the rotifers at once. 
Now rf*move the osmic acid as quickly as this can l)e done without taking up too 
many of the rotifei's (within a minute or two if iK)ssible), and wash si^veral times 
in distilled water. In thus fixing the rotifers in large numl>ers at once, it is usually 
imiM)ssible to draw off the osmic aci<l as soon as wouhl l)e lx\st, so that the animals 
l>ecome much blackened. But the blackening may l>e removed later with hydrogen 
peroxide. If the osmic aci<l has l)eiMi used at the right time usually a majority, or at 
least manv. of the rotifei-s will Ik? found to l)e fixed well extendtnl. l>ut as the time 
required for narcotization varies with difl'erent spe(*ies as well as with different 
individuals of the .same species, many of the animals will Im? found (*ontracte<l or 
with the structure j)artly obscured by deg(Mierative changes. With practice, how- 
ever, it will iMM'ome i)ossible to secure a sufficiently large j>ercentage kille<l in good 
condition to make the collection very valuable. 

Vi)v study of the loricate Rotifera it is advisable to kill some part of every col- 
lection directly by means of osmic acid, without previous narcotization, for in the 
loricate rotifers some of the most imix)rtant distinctive characters can l>est l)e seen 
in contracted specimens. 



ArUT wveral washing tlio rail eel. ion u ai-p prcwrvpii in 3 li> D per cent foi-iniiliii 
(3 l<i C parts ctimmi-rrial formalin to 100 jmrtu water). Tlioy can not In- jirt'sorveil 
in al<-4)hol without vausinf; o'xtonmvc slirinkatrR, reiKlcrinj; tlictii nwloss fur further 

These collections may later \to exaiiiiii<><1 under a h'-ns in onli'r to stiuly thi» 
rotifers Itoloiiftia^ to any family, ueniis, or species, nu<l Ihe s|ieciiiiens desii-ed picked 
ont hy means of a pipelle drawn Ui a capillary jxiint. The dilterent «|)ecies are 
sortod into difTerciit wat<'h i;lassi<s, and the l)lackcnin{i; due to the osmic acid ifl 
remove<l hy drawing olf inosi of the forinalin and adding; a few drops of hydi-ofp'n 
I»eroxidi' tor a few minntes. As «o<m as ilic dcsii-cd dcun'e of bleaching; is i-<>ache<l 
the liydrojren jw-roxidc is n>pla<>cd by formalin. The fornmliu should Ix- chaiifjiHl 
several t.iiiieH ami allowed 1o slaud Ncverai honi-n la-forv nioiuitinK the Npe<-iiiteus, 
otherwise buhblits of o\y;ren may appear niider the cover jila«» after it is sealed. 

SptM'inicns which have u«tt l>eeii in itsniio acid hnif! enouf^h to rei|uiri' blcacliinn 
are better in some ifspi-ds than those that have lieen bleached by the hydro;j;en 
IJcroxide, a.s the latter i-enioves the pifrment from the eye, as well us the blackeniuK 
dno to the osniie acid. 

The siM-einieus an' then monnled in hollo w- jtmii nd slides. The .slides should 
Iw thin and the concavities shallow, so Ihat it will Im- possible to use hifih powers of 
the mici-oseopo. The spe<'irneus are i ransferred to the concavities aton^ with some 
of the formalin, and covered with a cir-culiu- cover {jlass. It is best not to leave any 
bubbles of air Iwnealh the cover. Tlie sn|»erfliious fliii<l is withdrawn from ihe 
e«lKe of Ihe cover wilh a bit of filler pa|HT, ami the cover is then sealed. 

It is, of course, uccessnry to use some scidint: material that will not allow water 
to evapomte tlirough it. Mr. Housseh-t. the oiij;iiiiiior of this nieiii<Kl of mounting 
rotifers, reeommcnds the following; for scalinjr the mounts: After (i.xiny Ihe eover 
wilh a riiit; eomi>o.sed of a mixture of two-ihirds t;nni damar with one-lhiiil «old 
sisce, there are added two coats of pure shellac, followed by thrw or fiuir coats of 
jpild si/.e, alhiwinn twenly-fonr houi-s for ea<'h <'oal to dry. 

The followiuK accoiiiit of Ihe li,ill,iU.Uv is basi-d on the study of KH eollections, 
mtwlo as above, an<I represt'iitiiij: aln)nt half as many ditterent slalioiis. These 
collections wei-c mostly made almnt Ihe shores of l.aki- Kiie, durin;r Ihe suuiun-rs of 
ISftS, lKt<!) and 1!"H. while the writer was e<.nn.'<-led with Ihe biolojii<-al work on Ihe 
(Ji-eat Ijikes <-ai'ricd on by tlie I'nited States Fish Comndssion. The fulhiwia;! 
■-(■{rions were examined wilh HiM'cial thorouf;hncss: 

1. The i:ei;ioii aboul the ishuKls in Ihe western part ot" Lake Krie. 

•2. The soulh en- Ohio slioi-e of Luke Kri<-. in the ivyicTi kmiwn as Kiist ILn-bor, 
some distance fnnii San<lusky, Ohio. 

:j. The lake shore and river at llnnm, nhi.i. 

4. The ntjiion alM>ut Krie Harbor, Pennsylvania. Lii.ludiiif; the swamps and 
ptmils on I*n's(jiie Isle. 

.1. 1/uiK Toiot, on the <'amidiaa shon* of Irfikc Kric 

fi. Many collections have also liecn made almnt Ann .\rbor. .Mieli.. in |he iluroii 
River, and in a numlx-r of small stii-ams ari<l ponds in Ihe iieishb..rlioud. 

'Hiese eolliK-tions have been su|)plenienled by s|M'<-imcns ami miles fnndshed 
by a nnmlier of invesliiialoi-s in Kni-ojH', as mcniioiiefl in tin- ititriHluction. 



The RatiiUifJfe are Kotatoria, usually of small size, in which the cuticle of the 
bo<ly has become stiffened to form a sort of shell, railed a lori(»a. At llie anterior 
end is a ciliated area or corona, bv means of whi<»h the animal swims; this inav l>e 
retracted within the lorica. At the iM)sterior end is a small separate joint, known 
as the foot { /, figs. 1, -7, 4»'», etc.). To the foot are attached one or two bristle-like 
structures, which are called the toes. Tin* internal organs comprise an alimentary 
canal, nervous, muscular, cxcM-etorv, and repro<luctive systems, and certain mucus 
glands. In tin* following account these sets of organs will l>e taken up in order. 


(1) (ieneral farm. — The more usual form of tlie Iwxly in the Riithdidc^. is that 
of a cylinder, or long oval, fre(|uently curved. In some cases the Ixxly is much 
elongated, as in RatUdus elomjdfns (iosse (pi. Xll, fig. 1<>-), or DiureUa insiguis 
Herrick (pi. ll, fig. 15); in other eases it is short and plump, as in DiureUa porceUus 
Gosse (pi. II, figs. 10-21). In a few cas<^s ( RaftuJns Julus Jennings, pi. vii, figs. 05, G6; 
R. muJficriuis Kellicott, pi. vi, figs. 55-57) the bo<ly is broad and ovoid in form. 

A striking feature of the animals is their tendency to asymmetry in shape. 
This shows its^df in many ways. The body with the toes usually forms a curve, 
concave to the right, convex to the left (figs. 1, S, 1<;, 28, l^J, 1»5, 00, 102, etc.). The 
cur\'e is often not simple, l)ut is of such a nature that the body forms a segment 
of a spiral. This is perhaps liest seen in fig. 1, of DiureUa //r/r/.s MiiUer; it is a 
characteristic which is difficult to represent in a drawing, although often very 
noticeable in the animal itself. As will l)e seen later, the asymmetry shows it.self 
in the form and arrangement of many organs. 

(2) Larica — The body is covered with a hardened cuticula, known as the lorica. 
The lorica covers the body completely, being without openings at the sides, but it 
is oi>en anteriorly for the proj<M*rion of the corona, an<l posteriorly for the protrusion 
of the foot. The lorica is not so stiff an<i unyi<dding in the Raffulida as in many 
of the Kolifcra, usually i)erinitting considerable change of form. Compare, for 
example, the extende<l form of Rattuhis Jamjisttn Schrank (pi. viii, fig. 07) with the 
contracted form in the same species (])l. viil, fig. 7n). In some si)ecies the lorica is 
stifFer, not permitting such markcMl changes in shaiH*. 

Head-sheath. — The ant(»rior part of the lorica is usually s(»t off from the i*emain- 
der of the body by a slight constriction. This anterior portion, covering tin* head, 
ma}' 1m» known as the head-sheath (//. .s-., figs. 1, 3, 8, etc.). It presents a num]j<^r of 
interesting characteristic's, and .some that are very impoi'tant in classification. Only 
in Rattidus latus Jennings is it impossibh* to distinguish a head-sheath from the 
remain<ler of the lorica. 

The head-sheath fre<iuently has hingitudinal plaits, if they may l)e so designated, 
which serve for permitting the folding of the head-sheath when the head is retracted 
within the lorica. Th(\se are well s<*en in figs. 3, 4, 58, 50, and 02. These plaits 
seem to 1k» due to alternate longitudinal strips of hard, stiff material, and of soft, 
jielding cuticula. On the inner surface of the head-sheath aiv many fine transverse 
muscle fil)ers (sliown e.sp(»cially in figs. 58 and 50, pi. vi). When th<* head is drawn 
within the lorica, these longitudinal folds are lirought together b}' the yielding of the 


soft strips betwtvii Iheiii, and partly slip over one anotlier, so that the size of the 
head-sheath is greatly reduced and the anterior ojxMiinjr n(»arly or (^iiite elosed. It 
is possible to withdraw tlie head, at least partly, in most speeies without eausing the 
complete foldinj^: of the head-sheath; evidently a supplementary contraction of the 
fine transverse muscle Abel's is necessary to bring this a])out. 

In some species (notably Diurelln //f/r/.s Midler, pi. I, figs. 3, 4; Diurelltt rous- 
seleti Voigt, pi. iv, fig. M: Rattnhis mulfirrtniH Kellicott, pi. vi, fig. 5S; RnUiilus 
capucinus Wierz. &> Z^ich., fig. 51), and Raftnhfsrijlindricus Imhof, pi. vii, fig. Oi>) the 
head-sheath falls when contra^'ted into very regular folds. In /). iiyris Mftller, 
D. rou.sselefi Voigt, and />. int^n-media Stenroos, and perhaps in other species, the 
numl)er of these folds is nine. In some other species, as, for example, in R(tttulns 
(jracilis Tessin, pi. v, fig. 48, tlie folds are v(»ry irregular. In still other sin^cies no 
such folds are present, and the lorica may ivmain widely open when the head is 
retracted. This is the case, for (examples in Ratfidus srijiio (toshc, pi. v, fi)?. 52. 

On the anterior dorsal margin of the head-sheath there are in certain species 
of the Raiiuliihv a number of teeth. In Diurelhi rnusselefi Voigt theiv are nine 
well-marked teeth; in other si)ecies there are but one or two. Leaving out of con- 
sideration for the present the case of Diurella rousseh'ft\ wo may (dassify the teeth 
in other species into two categories: 

(a) In Raffulu.s mtiliicritiis Kellicott (pi. vi, tigs. 55 and 5S), RdfhiJtus mpuriniis 
Wierz. «fe Zach. (pi. vi, figs. 51)-61), and Raftulus ri/litHlrirus Imhof (pi. vii, fig. 62), 
there is a single nearly median j)rojection of the dorsjil lorica edge, exU^nding over 
the head. In Rafttilus rylindn'rus Indiof this is prolonged into a long hook, curved 
downward over the anterior oi>eniug of the lorica. In these cas<»s the tooth scmmus 
to l)e nearly or quite in the middle line. 

(b) In a numlH^r of other species there is eith(»r one toi^th {Raff itl us f/rncilis Tes- 
sin, figs. 45-48; Raftulus s<ipio (4osse, figs. 50-52; Diurella fhjris M(\ller, figs. 1, 3, 4; 
Diurella fenuior Gosse, figs. 7, S; Diurelht u^eheriy figs. 12-14 and 1KU117; Din- 
rella infermeflia ^tenrooH^ figs. 108, 10I>) — or two teeth (Raftulus louf/isffa Schrank, 
figs. 07-70; Diurella ijusiynis llerrick, figs. 15, 10: Diurella ])ar('^llus (io.sse, 
figs. 10, 20; Diurella sfylafa Kyferth, figs. 27-3n), whieh seem of a different cliar- 
act<*r. These lie distinctly to the right of the dorsal michlle line (so far as that can 
l>e define<l), ami form prolongations of one or both edges of the '\striated area" of 
the lorica,' described. Whcui there are two of these teeth they an> usually 
unequal in size, the right one Iwung longer. (Only in Diurella sttjlaf^f Eyferfh are 
they nearly or quite equal in length.) In most species they are merely short teeth, 
but in Raftulus loutjisefa Schrank and Diurella sfylafa Kyferth they are long spines. 
The position of these teeth on the right sich* is one of the markedly unsymmetrical 
charactei*s of the Raffulido'. A further ac(*ount of lhes(» teeth may best Im» deferred 
until the *' striated area'' has been described. 

Many of the 8i)ecies have no teeth at the* anterior edge of the lorica. The ante- 
rior opening of the lori<\*i is usually oval, with a slight notch near the v<»ntral middle 
line. In some few cas4»s the edge of the head-sheath j)rojects farther on the left side 
of the opening than on the right. This is notably the <*ase in Diurella ueheri n. sp. 
(pi. XIII, figs. 110-117); it is slightly so in Diurella ffuuifw (iosse an<l Diurella 
hrarhyura (tossc, and jx^rhaps in otlier species. 

In som(» cases thre<» or four or more t«»eth have Ixn^n descrilxMl by different 
authors at the ant<»rior edge of the lorica. In many cases this is due to the optical 


efTei»t of Die loii^tii<linal folils in the hea<l-slu»ath above desoribed or to the slight 
i*oiiiide<l projections of et*rtaiu parts of the head-sheath mentioned in the last para- 
jjraph. Sometimes the folds of the head-shealh project as sharp teeth. An example 
of this condition is fonnd in Diurelhi ntH.ssf^hfi ^'oigt, where th4»iv are nine of these 
teeth. These, however, an* of different (•haraet4»r from the one or two teeth which 1 
have descril)ed above. Th«»s<» latter arc structures to a <*prtain extent sui (jfiieris^ 
and I shall, as a rule, restrict the us<» of th<» term Ut^th in this (connection to them. 

Stenroos(l SOS) has described a new species. Mnst'KjiHfrca (Ruttnlus) rosfn^ which 
is said to have two long teeth or spines, like those of Baftnlus lomjisdit Schrank, at 
the ventral margin of the lorica. In other respects the animal resembles Raffulii.s 
lonijiseia Schrank. As this peculiar i)osiiion of the teeth is unknown in any other 
of the Rathdi(la\ and is entirely out of harmony with the struct un» an<l l)eliavior of 
the Rnthiluht in other respects (as will apiM*ar later), it s<NMns possible that then* 
was an error of ol>servation in this case. 

Hiriiited nrt^a^ R'uhjf. — One of the most pe<*uliar characteristic's of the Raihdithi 
is the pres(»nce on the lorica of a dorsal longitudiiml area, striateil transversely, 
which extends from the ant<»rior edge some distan<*e backward on the b<xly. This 
area shows the most varied <lifferentiati(ms in different species — in .some api>earing 
as a single high ridge, in otlu^rs as two ridges, in othei*s as a depression, while in 
still other cases there is no change in the surface of the lorica at this region except 
the transversa* striations. This |>eculiar area is so characteristic for the Raftulida^ 
and plays such a part in determining their forms, that it must Ix* tn'al4*d in full. 

The area is unsymmi*trical in position, usually In^ginning at the anterior margin 
of the lorica, to the right of the mid-dorsal line, an<l passing obliquely backward 
and towanl the left side. Its sitles are. as a rule, rather sliari)ly defintMl, fnujuently 
appearing as thickenings or ridges. This area shows in Raffuhis ^hmijulns (tossc a 
condition which will serve as a useful point of d(»i)arture for an undeiNtanding of the 
various differentiations which it undergOi\s in other species. In R. flotttjntHs Goase 
(pi. XII, fig. 1^2) the area begins at tht* anterior edge as a broa<l, shallow furrow, with 
well-marke<l sides. This furrow lies a litth' to the right of the position of the eye, as 
seen from alH)ve. From the sides of the furrow transverse striations j)ass t^)wanl 
its middle (and a little forwanl). The striations an* not (continuous from one side 
to the other, but me<»t in the middh* of th<* furrow in a sort of rliapln*. 

The furrow proper (*xtends backward for a distance only som<»whal gn^ater than 
the dianif*terof the lorica. Near its posterior end, in its middh* line, is situate<l tin* 
dorsal ant(*nna. Though the furrow or depn*ssion Indow th<* gen(*ral surfac** C4»ases 
at th«* point above indicated (shown at ./*, figs. 1<»2 and !<>•")), the striat4Ml aiva con- 
tinu4»s, with well-defined edges, for about one-thinl the l(*ngth of tin* lorica. 

In Riithdus lonyisehi S(*hrank (pi. viii, fig. 07) the striatcnl area is of very nearly 
the same character as in Raffuhis tlongafns (ioss(*, save that it exists as a depres- 
sion lhn)ugh()ut its entire IcMigth, reaching to the middle of tin* lorici^i. In this 
s|)eci4*s we hav<* another characteristic f(*atin*e added — th«* n*lation of thestriat^nl 
area to the two anterior te4*th or spin4*s. Thf^ hm ffeth an mnthnnitions of the 
thirkenfd fdyes (tf tite sfriaffd furron-. This appeal's to 1h* true in all sjM*ci4*s where 
the teeth exist. The tooth or spine which forms the continuation of the right 4*dg<» 
is much longer than tin* l«*ft one. 

What is the function of this striat4Ml area and what are the transvei-se striations 
which mark it? The striated furniw, as we find it in Rnftuhis hnifjisetn Schrank, 


beai*8 iiiudi ivscMnblance to one of the lonjjritndinal folds in the head-sliealh of 8U(*h 
species as NaUuhi.s raj^ucinti^s Wierz. it Zach. and RaUulufi tii nUirrinis Kellieol t (pi. vi, 
figs. 58 and 59); and these folds are cross-stria t<»d, jnst as in the case of the furrow. 
The striations in the folds of the head-sheath are evidently tine muscular hands, 
whicli have tlie office of bringing the fohls together when the liead is withdrawn. 

In the case of the dorsal striated ai*ea, it seems beyond question that the stria- 
tions are of the same nature — that they aiH» muscular bands. They are clearly not 
surface markings, but are internal bands. This is seen with esi)ecial ease in such 
forms as lintfuhi.s carinatus Lamarck and Rafixdus hicriMatus (xosse, in which the 
striated area rises in the form of one or two ridges. Moreover, the two e<lges of 
the furrow may l>e closely approximated, when the animal is strongly retracted, as 
in pi. VIII, fig. 70. When the head is extended the bases of the two teeth (on the 
opposite sides of the furrow) are a considerable distance a[)art; but when the 
animal is contracted to a maximum degree the two are almost in contact. 

The striated area therefore represents a longitudinal flexible portion of the 
lorica, permitting an increase or decrease in the ciriMimference of the bo<ly. The 
striations are muscle fibei's, bj' means of which the approxinmtion of the two sides is 
brought al)out. These fibers are attached at the middle and at. th<* two thickene<l 
edges of the area. 

In Rdffuhh^ mwosHs Stokes (pi. x, tig. S<)) the two edges of the striate<l area 
ai'e raised as pronounced ridges, leaving a broad and deep furrow l)etween them. 
The striations (muscle fibers) pass from the summit of the ridges t4) the lM)ltom of 
the furrow. Stokes (ISt)ri) states that he has seen the two ridges <lrawn toward each 
other, and I beli(*ve that I liave obs«»rved the same thing. 

In Rdfinlus hicrisffitus (tOssc the two edges of the area reach their highest 
development, rising as two v(»ry high i^rominent ridges with a liroad, deep furrow 
between them (pi. ix, figs. 77 and 7S). The muscles are groui^ed in pr(m<ninced 
bundles, which pass from near the summit of the ridges to the middle of the broad 
groov<^ between them. In a sqimrely side* view of the ridges the ends of the muscle 
bundles an* seen as irregular areas. 

In another series of species, of which Rattulus cfirinafus I^marck (pi. XI, figs. 
05, 117), Ratiuliis l()j)hot'ssns (tossc (figs. i>H, 09), and DiurelUi tiyris Miiller (fig. 1), 
lufiy be taken as types, only tlie riglit edge of the striat<Hl area is elevated into a 
ridge, the left not rising above the general surface of the bodj-. Thus a single ridge 
is pnxluced, having its edg«» toward the right, and sloping gradually to the left. 
The left edg«» of thc^ striated area may usually be recognized as a sharj), well-detintnl 
line, but not at all elevated. The muscle fibers run from the summit of the ridge 
(on the riglit) to the base of the ridge, at the left boundary of the ai'ea. The inter- 
ruption of the fibers in th(* middle of the area can usually still 1m> made out (though 
it is not indic'ated in all the figures). 

Thus we have produced the i)eculiar condition found in nmny of the RaHidithe 
and well shown in fig. 1 and fig. 95 (pi. xi) — a higli, sharp ridge passing cm the 
right side of the ])ody obliquely backward. Wh}' the right ridge should thus have 
develoi>ed rather than the left one we shall try t<) bring out in our gen<»ral discussion 
of the asymmetry of the RattuVuhv, 

In addition to the types already described the striated area is i)resent, in a (*on- 
sidera])le number of si)ecies, neither in the form of a well-defined ridge nor as a 


vvellMlelined ^move, but nuMoly as a flexible area with inarktKl trausversi* striatioiis. 
This is the case, for c^xauiple, in BaffuIiLS rfiilus Miiller (pi. xi, ligrs. 1(X), lol). In 
this organism the striated area is in some eases apparently swollen out to form a 
slight rounded ridj^e; in other eases it seems to lie at the general level of the loriea 
surface, while in still other specimens it seems to form a slight depression. It is 
probable that tln»se are functional dilTerences, <lue to the state of contraction or 
extensicm of the spcMMmen. Almost every intergradation is found, from the furrow 
of a. elonynffis (tossc to the high ridge of if. car hiatus Lamarck. In perhaps the 
majority of spe<»ies (<»si)ecially in Diunlhi) the striated area is merely slightly 
elevated at its right edg<% forming a low ridge, not conspicuous in most views. 

Th<» area in which tin* transverse striations can be seen usually passes from the 
anterior o<lge to the middle of the length of the bo<ly, or to a point souk* distance 
behin<l the mi<ldle. Tht* lidge formed by the eh^vation of the ar(»a sometimes con- 
tinues back farther than the striations, and may extend to the Ix^ginning of the 
f<x)t (as in MaUulus Joplioessus Gosse, pl. xi, figs. '.'><, lH')- 

Among the species which I have stu<li<Ml with care only Raffulu.s /f////.s- Jennings 
and jHM'haps Ratfahis mulfirrinis Kellic4>tt and RnttuJus capufiints Wiei^z. & Zach. 
show no sign of the striat(*d area. 

A word further should 1h» said about the rtdation of the striated area t4> the 
teeth or spines at the anterior e<lge of the lorica. Those of the second categtny 
mentioned on page 2S0 are formed as outgrowths of the thickened edges of the 
slriated area. \Vh(»re two teeth are i)resent both the edges project, that formed by 
tiie right i'i\]ii' being usually the longer. When only one tooth is pres<»nt it is 
form(»d by a projection of the right edge of the an»a. 

Th(» anterior i)rojections of the first category mentioned on page 2S0, found 
only in Raff ul us cylindrirus Imhof, Raff ul us cdpuflnus Wierz. S: Zach., an<l 
Rattulus niuJticrinis Kellicott, are formed in a som(»what different wa}'. The 
initial staL^e in the pi-oduction of sueh a proje<*tion is found in Raffulus eJotifjafus 
(iosse (pl. XII, fig. 102); the entire widtli of tlic* striat<Mi aiea proj(»cts at the anterior 
iH\^o as a rounded IoIm*. In Rfiffulu.s rajnai/ms Wierz. S: Zach. and 7f. niulfi- 
rrhiis Kellicott th(» projection has develop<Ml into a larg(» triangular t(H>th. In 
RaftuJuscijJ'nuhiriis Imhof (pl. vii, fig. iVl) tlie tip of this tooth has further developed 
into a l<»ng hook, curved down over tin* corona. Tlie tlir(*e si)ecit»s showing this 
jKM'uliar <lifTerentiation oc(Mipy a different position from most of the oth<»r species 
in many other respects also. 

(3) F*K)t. — 'i'he foot is a short, conical structure attached to th<» bo<ly at the 
j)ost(»rior end. The foot shows little variation in structure, excerpt in size arid form, 
being in some <'ases short and thick, in others shinier. In a few cases {Diurella 
jtorcfiJus (toss(», J), sulrafa Jennings, etc.), the ff)ot is very small, so as to be hardly 
recognizable as a separate structun*. In some of these cases it is usually held 
completely ret ra<*ted within the Ixxly. Som<»times the foot shows one or two faint 
annulations whicli hav«» at times been <lescribed as joints. 

The most peculiar thing alM)ut the foot in the RaftulitJji is its usually unsym- 
metrical attachment to the bo<ly. The joint Ix'tween the foot and the Iwxly is 
commonly oblique, extending farther ba<'k on the left (or left doi'sal) side than on 
the right. This is well shown in fig. S(; (pl. x), fig. UU (pl. xi), and fig. lo3 (pl. xll). 
In some cases th<» posterior e<lge of the lorica proj<»cts backward som«» <listance over 



the fiK)t on the left Hide, liut u<)t «ii l.lio rinlit. Tlio foot is thus atltw'IwMl Ut tlie loriea 
in such a way that it can 1>en<l to tlio ri};ht, but not to the loft. 

(I) Ti>e». — The U»» form porliapa the most peenlinr I'haractt'riatic of the 
Raffuliilft'. Mostof thcKotifprahav)? 1 wo short iKtst^rior ftp|>«n<Iages attwche*! to the 
foot, jilacixl Hi<l<! by side, and, like most i>aire<1 orgaim, .siniilar in fonn ami size. 
Itut in the RattuUfhi' wp find tlu' two toes in the majority of oaews unequal, 
sometimes excessively so, and no longer side by side. In some species one of the 
toes has almost ilisapi>earcd, while the other has liecome immensely developed, 
forming a straight rod as long as the body (in Jtafliihi.t njlhuirirus Iniliof. for 
example, pi, Vil, fig. (12). 

i.rj IHurrlla ttffrii, MOllur; ifrj /), .*jrf. 

.- Eyfw 


lite, nhowlTiK unulua] n>aartlnn of the right ti 

k I H. (■.(!*'" 

The steps in the sj-ries of changes Ijy which this is l>i-ought atmut may iw elearly 
followed by comparing llie toes of different species. In n few niK't-U^n {DlitreUa 
tkjrin Miiller, />. Kulfatii Jennings, D. hiienneih'a Stcnr<M«, etc.) the two toes are 
still equal, a.-< in i>ther rotifei-s. Oncttf tliese will serve l>est as a starting-point. 
We will select Diitnlhi tiijrix Miiller, whose tj)es are shown in text-ligure 1, at a. 

The to<'s fiirni two long, curved, iminled, spine-like ro<ls of equal size. At the 
base of each ai-e four small flattened spines (so-called sabstyles), which usually lie 
closely applied (o ihc base of the ti>cs. The use of these snbstyles was iwintcd out 
by J'late (18rt(;), and will Iw i-eadily appreciated when one of tlie Inibils of the 
animals is understoorj. The ptwtcrior jmrt of the iKxly e<intains two large glands 
(pi. I, ligs. ;i, 4, III. I/.), which secrete a qiianlily of mucus, which is store<l np in 



two large ^acs (figs, o, 4, ///. r.). These sacs open one at the base of each toe, and 
discharge the miieus out upon tlie surfa(*e of tlu* toe. Thence it trails behind the 
animal as a long thread, by means of which tin* i-olifer attaches itself to various 
external objects an<l hangs in the water, as a si)id<M' by its thread. The mucus 
passes (mt of the sa<! betwet^n the substyles and the main toes, and the four substyles 
serve to direct its course out ah>ng th<^ sui'faee of t